Fact-checked by Grok 2 weeks ago

Multi-label classification

Multi-label classification is a supervised paradigm in which each instance from a is assigned a of multiple, non-exclusive labels from a predefined label space, extending beyond traditional single-label that associates only one label per instance. This approach models the inherent complexity of real-world data where entities often exhibit multiple attributes or categories simultaneously, formalized as learning a h: \mathcal{X} \to 2^{\mathcal{Y}} from a training set D = \{ (x_i, Y_i) \mid x_i \in \mathcal{X}, Y_i \subseteq \mathcal{Y} \}, where \mathcal{X} is the instance space and \mathcal{Y} is the label space, with predictions typically obtained by thresholding the output of a real-valued . Originating from early research in multi-label text categorization in the early 2000s, the field has expanded significantly since the mid-2000s, driven by applications in diverse domains including multimedia annotation—such as tagging images with multiple objects or scenes—bioinformatics for gene function prediction, and for topic modeling in documents. Key challenges include the exponential growth of the output space (e.g., $2^{|\mathcal{Y}|} possible label combinations for |\mathcal{Y}| labels), label imbalance, and the need to capture dependencies among labels, which can be (independent per label), second-order (pairwise correlations), or higher-order (complex interrelations). Methodologically, multi-label classification strategies are broadly categorized into problem transformation techniques, which reduce the task to single-label or binary problems—such as Binary Relevance (BR), which trains an independent classifier for each label, or Classifier Chains (CC), which sequentially models label dependencies—and algorithm adaptation approaches, which modify base learners like k-nearest neighbors (e.g., ML-kNN) or decision trees to directly handle multiple outputs. Recent advancements leverage architectures, including convolutional neural networks (CNNs) for visual data, recurrent neural networks (RNNs) and transformers for sequential data, and graph neural networks (GNNs) for explicit label correlation modeling, addressing scalability in extreme multi-label settings with millions of labels and improving performance through techniques like attention mechanisms and loss functions tailored for imbalance (e.g., focal loss variants). These developments have enhanced evaluation metrics beyond simple accuracy, incorporating measures like Hamming loss, subset accuracy, and ranking-based metrics such as mean average precision (mAP) to assess both label-wise and instance-wise performance. areas continue to focus on weakly supervised scenarios with partial labels, efficient handling of high-dimensional features, and robust across imbalanced or noisy datasets, as well as integration with large language models and few-shot/ paradigms.

Fundamentals

Definition and Motivation

Multi-label classification is a supervised learning paradigm in which each instance in a dataset is associated with a subset of labels from a predefined finite set of possible labels, allowing for multiple non-exclusive assignments to a single instance. Unlike traditional single-label classification, where an instance is assigned exactly one label, multi-label classification recognizes that real-world data often exhibits overlapping categories, enabling more nuanced representations of complex phenomena. The primary motivation for multi-label classification stems from its ability to model the inherent multiplicity in many practical applications, where forcing a single label would oversimplify the underlying structure and lead to information loss. For instance, in image annotation, a single photograph might simultaneously depict a "beach," "sunset," and "people," requiring all relevant tags for accurate retrieval and organization. Similarly, in text categorization—a domain that heavily influenced its development—a news article could pertain to "politics," "economy," and "international" topics, reflecting interconnected themes that single-label approaches cannot capture effectively. This flexibility makes multi-label methods essential for tasks like protein function prediction, music genre tagging, and medical diagnosis, where instances naturally belong to multiple classes. Multi-label classification emerged in the late , largely motivated by challenges in text categorization, where documents often span multiple topics within hierarchical or flat label sets. Early work addressed these needs through probabilistic models, such as McCallum's (1999) trained via expectation-maximization for assigning multiple categories to texts. A seminal advancement came with Schapire and Singer's (1999) introduction of the .MH algorithm, which adapted boosting techniques to handle multi-label problems efficiently, particularly for large-scale text . These developments laid the for subsequent research, highlighting the paradigm's utility in managing label dependencies and non-exclusivity.

Formal Problem Statement

In multi-label classification, the input consists of an instance space \mathcal{X}, typically comprising feature vectors x \in \mathbb{R}^d where d denotes the dimensionality, and a finite set of q possible labels forming the label space \mathcal{Y} = \{1, 2, \dots, q\}. For each instance x \in \mathcal{X}, the output is a Y \subseteq \mathcal{Y} of relevant labels, which is commonly encoded as a vector y \in \{0,1\}^q with y_j = 1 if label j is relevant to x and y_j = 0 otherwise. The learning task involves a training \mathcal{D} = \{(x_i, y_i)\}_{i=1}^n, where each pair consists of an instance x_i \in \mathcal{X} and its corresponding label y_i \in \{0,1\}^q. The goal is to induce a predictor f: \mathcal{X} \to [0,1]^q that outputs a of confidence scores or probabilities for each , from which predictions \hat{y} \in \{0,1\}^q can be obtained by applying a (typically 0.5) to each component. The objective is to minimize an empirical over the training data, such as the Hamming loss, defined as \frac{1}{nq} \sum_{i=1}^n \sum_{j=1}^q \mathbb{I}(\hat{y}_{ij} \neq y_{ij}), where \mathbb{I}(\cdot) is the that equals 1 if the argument is true and 0 otherwise, \hat{y}_{ij} is the j-th component of the predicted for instance i, and y_{ij} is the corresponding true label. This loss quantifies the fraction of label-wise prediction errors across all instances and labels. Unlike single-label settings, multi-label classification assumes that labels may exhibit dependencies among themselves and imposes no , allowing an instance to be associated with any non-empty subset of \mathcal{Y}, including the full set or none at all (though the is sometimes handled separately in practice).

Binary and Multi-class Classification

is a fundamental task where each instance is assigned to exactly one of two mutually exclusive classes, often represented as y \in \{0,1\}, with a separating the classes exclusively. This setup is common in scenarios requiring simple yes/no decisions, such as detection, where an incoming message is categorized as either or legitimate (). In such cases, the model learns to distinguish patterns indicative of one class versus the other, assuming no overlap in labeling. Multi-class classification generalizes to scenarios with more than two classes, where each instance receives exactly one label from a set of q > 2 mutually exclusive categories, denoted as y \in \{1, \dots, q\}. This problem is frequently addressed through decomposition strategies, such as one-vs-all ( q binary classifiers, each treating one class against all others) or one-vs-one ( \binom{q}{2} classifiers for pairwise distinctions). Examples include image recognition tasks like classifying fruits into apples, oranges, or bananas, where classes are disjoint and only a single assignment is made per instance. In contrast, multi-label classification—as outlined in the formal —permits instances to belong to multiple classes simultaneously, allowing overlapping labels and resulting in an output space of $2^q possible label subsets, far exceeding the q possibilities in multi-class settings and leading to a in complexity. For instance, while or multi-class approaches might classify a article into a single topic like "," multi-label classification could assign multiple tags such as "," "," and "" to the same article. This non-exclusive labeling introduces unique challenges, including severe label imbalances (where some label combinations are rare or absent) and the need to model inter-label correlations, which single-label methods like and multi-class classification typically overlook by treating decisions independently.

Multi-instance and Other Variants

Multi-instance learning (MIL) is a supervised learning paradigm where training data consists of bags of instances, each bag associated with a single label, typically binary. A bag is labeled positive if at least one of its instances is responsible for the positive classification, while negative bags contain only negative instances; the individual instance labels are unobserved. This formulation was introduced to address problems where precise instance-level labeling is infeasible or costly, such as in drug activity prediction, where a molecule (bag) is active if any of its low-energy conformations (instances) binds to a target protein. In contrast to multi-label classification, which assigns multiple independent labels directly to each instance based on its features, MIL emphasizes aggregation and ambiguity at the bag level, with the goal of identifying the "witness" instances that determine the bag's label without per-instance annotations. While extensions like multi-instance multi-label learning combine the two by associating multiple labels with bags of instances, standard MIL differs fundamentally by focusing on bag-level decisions rather than per-instance multi-labeling. Hierarchical classification extends multi-label settings by organizing labels into a structured , such as a or (DAG), where labels represent parent-child relationships and predictions must respect semantic dependencies, such as assigning a only if the parent category is predicted. For instance, in protein function , labels might form a , allowing an instance to receive multiple labels at various levels while enforcing consistency (e.g., no leaf label without its ancestors). This overlaps with multi-label classification, as hierarchical problems are often multi-label in nature but incorporate structural constraints to improve accuracy and interpretability; multi-label methods can be adapted for hierarchies via , but pure multi-label treats labels as flat and independent. Extreme multi-label classification addresses scenarios where the label space is vast, typically exceeding 10,000 labels, requiring efficient algorithms to predict sparse, relevant subsets from millions of possible tags, as in tagging or product recommendation systems. Unlike standard , which handles moderate label counts (e.g., tens to hundreds), extreme variants prioritize scalability and over exhaustive prediction, often using embedding-based approximations to manage . Ordinal classification, another variant, imposes a on labels, treating them as ranked categories (e.g., low/medium/high severity in ) rather than independent or equally spaced, which contrasts with multi-label's assumption of unrelated labels; while multi-label can incorporate ordinal aspects through graded memberships, standard ordinal focuses on single-output with ordered thresholds. These variants highlight structural extensions of problems, with multi-label serving as a foundational framework that can integrate hierarchical or ordinal elements for domain-specific refinements.

Problem Transformation Methods

Binary Decomposition Approaches

Binary decomposition approaches transform the multi-label classification problem into multiple independent tasks, one for each label, thereby avoiding the need to model label dependencies explicitly. The most straightforward and widely adopted method in this category is Binary Relevance (BR), which trains a separate binary classifier for each of the q labels in the label set. In BR, given a training set of instances \{( \mathbf{x}_i, \mathbf{y}_i ) \mid i = 1, \dots, n \}, where \mathbf{x}_i \in \mathbb{R}^d is the vector and \mathbf{y}_i \in \{0,1\}^q is the vector, a classifier f_j: \mathbb{R}^d \to [0,1] is learned for each j = 1, \dots, q to estimate the probability P(y_j = 1 \mid \mathbf{x}), typically by treating the j-th column as the in a standard problem. For a new instance \mathbf{x}, the predicted set is obtained by applying a threshold (often 0.5) to each f_j(\mathbf{x}), yielding \hat{\mathbf{y}} where \hat{y}_j = 1 if f_j(\mathbf{x}) \geq 0.5, else 0. This approach inherently minimizes the Hamming loss for each individual classifier, as the overall Hamming loss in multi-label classification decomposes into the sum of per- losses under the independent assumption. BR offers several advantages, including simplicity in implementation, the ability to use any off-the-shelf binary classifier, and parallelizability during training since classifiers are independent. However, its primary limitation is the complete disregard for label correlations, which can lead to suboptimal performance when labels are dependent, as it cannot exploit co-occurrence patterns to improve predictions. A practical example of binary decomposition is in image tagging tasks, where separate binary classifiers are trained for tags like "cat," "dog," and "outdoor"; an input image is then assigned all tags whose classifiers output probabilities above the threshold, enabling efficient annotation without assuming mutual exclusivity.

Multi-class Transformation Approaches

Multi-class transformation approaches in multi-label classification convert the problem into a single multi-class classification task by mapping each possible subset of labels to a unique class. This strategy inherently models dependencies and correlations among labels, as the classifier learns to predict entire label combinations rather than treating labels independently. Unlike binary decomposition methods, which train separate classifiers for each label, multi-class transformations consolidate the prediction into one model that outputs a label subset directly. The Label Powerset (LP) method is a foundational example of this approach. It transforms the multi-label dataset by assigning each distinct label subset observed in the training data to a unique class, resulting in a multi-class problem with up to $2^q classes, where q is the number of possible labels. A standard multi-class classifier, such as a or , is then trained on this transformed data. To mitigate the exponential growth in class count for large q, LP often prunes rare or unobserved subsets during training. Prediction under LP involves selecting the label subset with the highest probability given an input instance: \hat{Y} = \arg\max_{Y \subseteq \mathcal{Y}} P(Y \mid x) where \mathcal{Y} is the set of all labels, x is the input feature vector, and \hat{Y} is the predicted subset from which individual labels are extracted. This formulation allows the model to capture label correlations by learning joint probabilities over subsets. A key advantage of LP is its ability to explicitly account for label dependencies, leading to improved performance in domains where co-occurrences are meaningful, though it suffers from high and potential when q is large due to the explosion in the number of classes. To address these limitations, variants like , also known as Pruned Sets (), reduce the effective number of classes by eliminating infrequent label subsets—typically those appearing fewer than a (e.g., three) times in the training data—while preserving dominant combinations. This enhances and without fully sacrificing modeling. PLP has shown empirical gains in accuracy on datasets with sparse label distributions. In practice, LP and its variants are applied in text classification tasks, such as tagging documents with multiple categories (e.g., combining "sports" and "news" into a single class for training), where label subsets represent thematic overlaps and enable more coherent predictions.

Ensemble-based Methods

Ensemble-based methods in multi-label classification combine multiple instances of problem transformation approaches, such as binary relevance or label powerset, to enhance predictive performance by leveraging the strengths of ensembles like reduced variance and improved robustness to label dependencies. These techniques build on basic transformations by training several models and aggregating their outputs, often through averaging or voting, to mitigate limitations like error propagation in single chains or computational intractability in large label spaces. One prominent ensemble method is the Ensemble of Classifier Chains (ECC), which addresses the sensitivity of single classifier chains to label ordering by averaging predictions across multiple chains with randomized orders. In a single classifier chain, for a given chain order \pi, the j-th binary classifier f_j predicts the relevance of label \pi(j) based on the input features x and the predictions of preceding labels: f_j(x, y_{\pi(1)}, \dots, y_{\pi(j-1)}). The ECC trains an ensemble of m such chains, each on bootstrap samples of the training data, and computes the final confidence for label j as the average over all chains: \hat{w}_j = \frac{1}{m} \sum_{k=1}^m \hat{y}_{j,k}, where \hat{y}_{j,k} is the prediction from the k-th chain; labels are then thresholded at 0.5 to obtain binary predictions. This approach was introduced by Read et al. in 2011, demonstrating superior performance over binary relevance on datasets with correlated labels by capturing inter-label dependencies while averaging out order-specific errors. Another key ensemble is RAndom k-labELsets (RAkEL), which ensembles Label Powerset classifiers applied to random subsets of labels to approximate label powerset transformations more efficiently for large label spaces. RAkEL partitions the full label set into m random subsets of size k (typically k \ll M, where M is the total number of labels) and trains a label powerset classifier on each subset, treating the subset as a multi-class problem; predictions are combined via or averaging across subsets. Variants include RAkELd, using disjoint subsets for balanced coverage, and RAkELo, using overlapping subsets to better capture correlations through . Proposed by Tsoumakas et al. in 2010, RAkEL improves upon standalone label powerset by reducing in M and enhancing accuracy on skewed distributions, as shown in experiments on benchmark datasets like and . Other ensembles include bagging or boosting variants of binary relevance, such as Ensemble Binary Relevance (EBR), where multiple independent binary classifiers per label are trained on bootstrapped samples and averaged to stabilize predictions. For instance, boosting multiple binary relevance models has been applied to robust text tagging tasks, iteratively weighting misclassified instances to improve handling of imbalanced labels. These methods balance the assumption of label independence in with correlation modeling in or approaches, often yielding higher accuracy than single transformations on correlated datasets. However, they incur higher computational costs due to multiple models, scaling linearly with ensemble size but potentially requiring for very large datasets.

Direct Algorithm Adaptations

Instance-based and Nearest Neighbor Methods

Instance-based and nearest neighbor methods adapt algorithms, such as k-nearest neighbors (kNN), to directly handle multi-label outputs without an explicit training phase, relying instead on the training data during prediction. These methods identify the k most similar instances to a query instance based on feature space distances, then aggregate label information from those neighbors to predict multiple labels. A seminal approach in this category is the Multi-label k-Nearest Neighbors (ML-kNN) algorithm, which extends traditional kNN by estimating the of each label using on the neighbors' label memberships. In ML-kNN, for a query instance x, the k nearest neighbors N_k(x) are found using a standard distance metric like in the feature space. The posterior probability that label y_j is relevant (y_j = 1) is then computed as a smoothed estimate incorporating the proportion of neighbors with that label and the \hat{\mu}_j: P(y_j = 1 \mid x) = \frac{\sum_{i \in N_k(x)} y_{ij} + m \hat{\mu}_j}{k + 2m}, where m is the Laplace smoothing parameter (typically m=1) to avoid zero probabilities, and \hat{\mu}_j is the prior probability of label j estimated from the training set. The label set is determined by thresholding these probabilities, often at 0.5, using the maximum principle. This approach has demonstrated competitive performance on benchmark datasets, such as the Yeast gene functional task, outperforming methods like BoosTexter in terms of Hamming loss (0.194 vs. 0.220). Other adaptations incorporate label-based distance metrics to better capture correlations between instances via their label sets, such as the , which measures the proportion of differing labels between two instances: d_H(Y_1, Y_2) = \frac{1}{q} \sum_{j=1}^q |y_{1j} - y_{2j}|, where q is the number of labels. These metrics can be combined with feature distances (e.g., via weighted sums) to refine neighbor selection, improving prediction in domains with sparse or correlated labels, as explored in extensions like nearest labelset methods. The advantages of these methods include their simplicity, interpretability through neighbor inspection, and lack of a separate training phase, making them suitable for dynamic datasets. However, they are sensitive to the choice of k and distance metric, with computational cost scaling as O(N) per query in large training sets N, and performance degrading in high-dimensional spaces due to the curse of dimensionality. A practical example is music or movie recommendation, where user preferences are multi-label (e.g., multiple genres like , and ). For a new x, ML-kNN finds similar users' profiles, estimates probabilities for each based on their assigned , and recommends a set of genres exceeding the , enabling personalized suggestions without assuming label independence.

Tree and Rule-based Adaptations

Tree and rule-based adaptations extend classical and induction algorithms to handle multi-label classification by modifying splitting criteria, predictions, and generation processes to account for multiple s per instance. These methods retain the interpretability of their single-label counterparts while adapting to label sets, often through probabilistic or threshold-based mechanisms at decision nodes and leaves. Multi-label decision trees (MLDT) represent a key adaptation where the is built recursively by selecting features that maximize information gain based on a multi-label measure. At each internal node, the splitting criterion evaluates the reduction in uncertainty over the joint label distribution, defined as H(Y) = -\sum_{Y' \subseteq \mathcal{Y}} p(Y') \log p(Y'), where \mathcal{Y} is the label space and p(Y') is the probability of subset Y'; however, due to exponential complexity, approximations like summed binary entropies are often used in practice. At the leaves, the predicted label set is determined via majority vote over the labels assigned to instances reaching that leaf, selecting labels present in more than half of the examples. This approach builds an explicit hierarchical model during , contrasting with lazy methods like instance-based learners that defer computation. For rule-based adaptations, algorithms like RIPPER (Repeated Incremental Pruning to Produce Error Reduction) are extended to multi-label settings through iterative separate-and-conquer strategies that generate rules predicting label subsets or incorporating label dependencies. In one adaptation, rules are learned sequentially for each label, with predictions from prior rules fed back as features for subsequent ones, enabling a decision list that covers positive and negative label outcomes (e.g., SeCo^\pm variants). This produces compact, interpretable rule sets where each rule specifies conditions for assigning a combination of labels. The original RIPPER framework supports to reduce , which is preserved in multi-label versions to balance accuracy and simplicity. These adaptations offer strong interpretability, allowing domain experts to inspect decision paths or rules for insights into label relationships, as seen in applications like tagging medical symptoms with multiple diseases—e.g., a tree might split on patient age and symptom severity to predict co-occurring conditions such as and at leaves. However, without explicit extensions for label correlations, they may underperform on datasets with strong inter-label dependencies, as splits and rules often assume . Training efficiency remains a strength, with linear in feature dimensionality for trees.

Neural Network and Deep Learning Adaptations

Neural networks and models adapt to multi-label classification by modifying the output layer to produce independent predictions for each label, typically using a activation function to map scores to probabilities between 0 and 1 for each of the q labels. The binary cross-entropy (BCE) loss is then applied independently to each label, formulated as \mathcal{L} = -\sum_{j=1}^q \left[ y_j \log \hat{y}_j + (1 - y_j) \log (1 - \hat{y}_j) \right], where y_j is the true binary label for the j-th class and \hat{y}_j is the predicted probability, allowing the model to handle non-exclusive labels without assuming . This approach treats the problem as q parallel binary classifications, enabling end-to-end training via on architectures like convolutional neural networks (CNNs) or recurrent neural networks (RNNs). To capture label correlations, attention mechanisms are integrated into deep models, enhancing the representation of dependencies between labels during training. For instance, in text-based multi-label tasks, transformer-based architectures employ self-attention to weigh relevant tokens and labels jointly, improving performance on datasets with interrelated categories. A notable example is AttentionXML, which uses a multi-label attention mechanism over raw text and a label tree structure to efficiently model correlations in extreme settings, achieving up to 17% higher precision than baselines on large-scale datasets like Wiki-500K. These adaptations leverage the transformer's ability to process sequences in parallel, making them suitable for capturing higher-order interactions beyond independent assumptions. For extreme multi-label classification, where q can reach millions, embedding-based methods like XML-CNN address scalability by using CNNs with dynamic max pooling on word embeddings, followed by a layer to project to the label space and BCE loss for prediction. This enables handling of vast label sets on datasets like with 500,000 labels, though subsequent analyses have identified implementation issues affecting reported performance. More recent advancements include fine-tuned models, such as adaptations for multi-label tasks, and parameter-efficient techniques like adapters or low-rank adaptations () to scale direct neural methods while reducing computational demands, as demonstrated in benchmarks up to 2024. adaptations excel at capturing complex, non-linear dependencies in high-dimensional data, offering superior accuracy over classical methods in domains like and text tagging. However, they require large amounts of for effective training and can act as black boxes, complicating interpretability. A representative application is CNN-based multi-label image classification on the COCO , which features 80 object categories across 123,000 images. Models like CNN-RNN combine convolutional feature extraction with recurrent layers for label dependencies, using BCE loss to achieve mean average precision improvements over baseline CNNs by modeling co-occurring objects.

Learning Paradigms

First-Order and Independent Assumptions

In multi-label classification, the first-order paradigm treats the labels as independent, optimizing separate loss functions for each label without considering interdependencies among them. This approach simplifies the problem by decomposing it into individual tasks, where the prediction for each label is made in isolation based on the input features. Seminal methods adopting this paradigm include , which trains one binary classifier per label, and ML-kNN, a adaptation of k-nearest neighbors that estimates label probabilities independently. The core independent assumption underlying this paradigm posits that the joint probability of the set Y given an instance x factors into the product of individual label probabilities: P(Y \mid x) = \prod_{j=1}^q P(y_j \mid x) where q is the number of labels and each y_j is a indicator for the j-th label. This assumption, analogous to the naive Bayes independence in single-label settings, enables efficient computation but holds only approximately in practice, as labels often exhibit given the features. Despite its simplicity, the first-order paradigm dominated early multi-label research before 2005, appearing in applications such as scene classification and text categorization where label correlations were initially overlooked. For instance, in tag prediction for images or documents, independent binary classifiers might predict tags like "beach" or "sunset" separately, potentially missing that "beach" and "ocean" frequently co-occur. A key limitation is its inability to capture label co-occurrence patterns, which can lead to suboptimal performance on datasets with strong dependencies, such as biological gene functions or medical diagnoses. Many direct algorithm adaptations, including instance-based methods, build upon this paradigm for its computational tractability.

Higher-Order Label Correlation Modeling

In multi-label classification, learning paradigms for handling label correlations are classified based on the order of dependencies modeled. approaches assume label , treating each label separately. Second-order methods incorporate pairwise correlations, such as P(y_j, y_k | x), where y_j and y_k are labels and x is the instance. Higher-order paradigms extend this to full joint distributions over label subsets, capturing complex interdependencies beyond pairs. Second-order modeling focuses on pairwise label correlations to improve prediction accuracy without excessive computational overhead. A seminal approach is the Classifier Chains (CC) method, which builds a chain of binary classifiers where each subsequent classifier incorporates predictions from previous ones as additional features, thereby modeling sequential pairwise dependencies. This paradigm enhances performance on datasets with moderate correlations while remaining scalable. Higher-order correlation modeling addresses limitations of pairwise methods by estimating probabilities over larger label subsets, such as full joint distributions P(\mathbf{y} | x), where \mathbf{y} denotes the label vector. Probabilistic Classifier Chains (PCC) extend CC probabilistically, enabling Bayes optimal predictions by approximating the joint distribution through factorization and efficient inference techniques like . Bayesian networks provide another framework, representing label dependencies as a that encodes conditional independencies and facilitates joint probability computation via structure learning and inference algorithms. These methods are particularly effective for datasets with intricate, non-pairwise correlations. Higher-order paradigms offer advantages in accuracy for tasks with strongly correlated labels, such as video tagging where "sports" and "outdoor" frequently co-occur, allowing models to leverage joint patterns for more coherent predictions. However, they introduce drawbacks including higher due to in label combinations and challenges in accurate from limited .

Evaluation Metrics

Threshold-based Metrics

Threshold-based metrics in multi-label classification evaluate the performance of models by first converting continuous output scores for each into predictions using a predefined , typically transforming the problem into multiple tasks for assessment. These metrics focus on the accuracy of the resulting assignments across instances and labels, providing measures that align with the nature of predictions in applications like text categorization or image annotation. Common include 0.5 for symmetric score distributions, though optimization via validation sets or methods like SCut can adjust per-label to maximize specific evaluation criteria. One fundamental threshold-based metric is the Hamming loss, which quantifies the fraction of labels that are incorrectly predicted across all instance-label pairs. It is formally defined as \text{Hamming Loss} = \frac{1}{nq} \sum_{i=1}^n \sum_{j=1}^q \mathbb{I}(\hat{y}_{ij} \neq y_{ij}), where n is the number of instances, q is the number of labels, y_{ij} is the true binary label for instance i and label j, \hat{y}_{ij} is the predicted binary label, and \mathbb{I}(\cdot) is the indicator function. Lower values indicate better performance, with a perfect prediction yielding zero loss; this metric tolerates partial correctness, making it suitable for scenarios where exact label sets are not required. Another key metric is the exact match ratio, also known as subset accuracy, which measures the proportion of instances where the entire predicted label set matches the true label set precisely. It is given by \text{Exact Match} = \frac{1}{n} \sum_{i=1}^n \mathbb{I}(\hat{Y}_i = Y_i), where Y_i and \hat{Y}_i denote the true and predicted label sets for instance i, respectively. This metric is stricter than Hamming loss, penalizing any deviation in the label set and thus emphasizing the importance of complete accuracy, though it can be overly harsh in domains with sparse or variable label densities. Precision, recall, and F1-score can also be adapted in threshold-based evaluations, either at the macro level (averaging per-label scores) or micro level (aggregating contributions globally across all labels). F1 averages the F1-score computed independently for each label, treating them equally regardless of frequency, while micro F1 pools true positives, false positives, and false negatives across labels before computing , which favors frequent labels. These variants provide balanced assessments of prediction quality, with macro F1 useful for imbalanced label distributions and micro F1 for overall error rates. In practice, such as music tagging where tracks may receive multiple or labels, Hamming loss effectively evaluates partial matches by rewarding systems that correctly identify some but not all tags, allowing for nuanced performance insights in recommendation tasks.

Ranking-based Metrics

Ranking-based metrics assess the quality of label predictions by considering the ordering or produced by a model's scores, rather than requiring a fixed to convert scores to decisions. These metrics are particularly valuable in multi-label settings where partial rankings can capture nuanced , emphasizing how well true labels are positioned relative to irrelevant ones in the predicted list. They differ from threshold-based metrics by focusing on ordinal relationships and partial credit for correct ordering, making them suitable for evaluating models that output continuous scores. One-error is a fundamental ranking-based metric that quantifies the proportion of instances where the highest-ranked label is incorrect, i.e., not part of the true label set. For a dataset of n instances, it is formally defined as \frac{1}{n} \sum_{i=1}^n \mathbb{I} \left( \arg\max_j \hat{y}_{ij} \notin Y_i \right), where \hat{y}_{ij} represents the predicted score for label j on instance i, Y_i is the ground-truth label set for instance i, and \mathbb{I}(\cdot) is the indicator function that equals 1 if the condition holds and 0 otherwise. This metric, introduced by Schapire and Singer in their work on boosting for text , serves as a direct analog to classification in single-label problems but adapted to prioritize the top prediction in multi-label contexts. Coverage evaluates the extent to which true are clustered toward the top of the list by measuring the average number of steps required to include all relevant labels when traversing the from highest to lowest score. It is calculated as \frac{1}{n} \sum_{i=1}^n \left( \max_{j: y_{ij}=1} r_{ij} - 1 \right), where r_{ij} is the position of label j for instance i (with being the highest score), and y_{ij}=1 indicates that label j belongs to Y_i. A lower coverage value signifies superior , as it implies true labels occupy higher on average. Schapire and Singer proposed this to assess the of relevant label placements in ranked outputs, highlighting models that minimize the span needed to achieve full . Average precision adapts the precision-at-rank concept from to multi-label evaluation, computing the mean precision achieved at the rank positions of each true label across the dataset. For each instance i, it averages the ratio of true labels among the labels ranked above (including) each true label's position, then takes the instance-wise mean. This yields a score between 0 and 1, where higher values indicate that true labels are not only highly ranked but also densely packed without many false positives interspersed. Originating from Schapire and Singer's framework for multi-label in tasks, average precision is favored for its robustness to varying label densities and its emphasis on early precision in the ranking. These ranking-based metrics find prominent use in domains where selecting and ordering a small number of top predictions is prioritized over exhaustive , such as generating search suggestions or personalized feeds, where depends on the of initial recommendations. For instance, in a movie recommendation system suggesting top-3 genres, ranking loss—a related measure penalizing pairwise inversions between true and false labels—complements one-error and coverage by ensuring true genres like "" and "" outrank irrelevant ones like "romance," thereby improving the utility of the ranked output.

Advanced Topics

Multi-label Stream Classification

Multi-label stream classification addresses the task of assigning multiple labels to instances that arrive continuously in an unbounded , where data is processed incrementally without storing the entire history. Unlike batch multi-label classification, this paradigm handles infinite sequences of examples, often exhibiting concept drift—abrupt or gradual shifts in the underlying label distributions or correlations—requiring models to adapt in while maintaining bounded memory usage. Key adaptations extend static methods to streaming settings. One prominent approach is the multi-label Hoeffding tree, which builds decision trees incrementally using the Hoeffding bound to select splits with high confidence after observing sufficient examples, while incorporating multi-label predictors at the leaves to account for label dependencies. For instance, the Scalable Multi-label Hoeffding Tree employs binary relevance (BR) transformations at leaf nodes, where each label is treated as an independent binary classification problem solved via Hoeffding classifiers, combined with pruned sets to filter unlikely label combinations. To handle forgetting and drift, these trees use adaptive mechanisms like sliding windows or explicit replacement of outdated subtrees. A more recent variant, the Multi-Label Hoeffding Adaptive Tree (MLHAT), enhances the classic Hoeffding Adaptive Tree (HAT) by integrating label co-occurrences into split criteria and dynamically updating leaf learners for evolving correlations, demonstrating superior performance across diverse stream benchmarks. Online BR methods further adapt by training separate Hoeffding classifiers per label with forgetting strategies, such as fixed-size buffers or drift-triggered resets, to discard obsolete examples and prevent memory explosion. Challenges in multi-label stream classification include enforcing limited for infinite data, ensuring predictions under high , and detecting subtle drifts in label shifts or correlations without supervision. Drift detection algorithms like ADWIN (Adaptive Windowing) address these by maintaining a variable-length window of recent examples and applying the Hoeffding bound to identify statistically significant changes in accuracy or distribution: specifically, it compares sub-windows' means \bar{x}_0 and \bar{x}_1 over lengths n_0 and n_1, flagging drift if |\bar{x}_0 - \bar{x}_1| > \epsilon, where \epsilon = \sqrt{\frac{\ln(1/\delta)}{2n}} for confidence \delta, allowing localized adaptation for label-specific shifts. This enables mechanisms like affected model components while preserving stable parts. A practical example is hashtag prediction on platforms, where tweets or posts arrive as and require assigning multiple evolving hashtags () amid trending shifts, such as sudden topics; hierarchical multi-label models adapted for can process these incrementally, updating hierarchies to capture emerging correlations like topic co-occurrences. Post-2020 advances include online models that integrate architectures for , such as Deep Streaming Label Learning, which uses a to map emerging labels to historical ones via mechanisms, enabling adaptation to label drifts in high-dimensional settings like text . Additionally, transformer-based approaches like those in concurrent multi-label prediction for event employ conditional mixtures of distributions to model label dependencies online, outperforming traditional trees on dynamic benchmarks.

Challenges and Future Directions

One of the primary challenges in multi-label classification arises from issues, particularly in scenarios involving an extreme number of labels, such as q > 10^5, which is common in large-scale recommendation systems where millions of items or tags must be predicted. The high dimensionality of the label space leads to significant computational bottlenecks, including constraints and inefficiencies, as traditional methods struggle with the sparsity of label co-occurrences and the inability to train individual classifiers per . For instance, in recommendation tasks, the vast label space exacerbates sparsity, where most instances are associated with only a few labels, making it intractable to capture rare associations without specialized approximations. Label and imbalance further complicate multi-label classification, especially with partial labels—where only a subset of relevant labels are provided—and long-tail distributions that favor frequent labels at the expense of rare ones. Partial labeling introduces in training, as models must infer missing associations while handling noisy annotations that degrade performance in imbalanced settings. Long-tail distributions amplify this by causing models to underperform on minority labels, leading to biased predictions that overlook tail-end categories critical in domains like or content tagging. Approaches addressing these issues, such as correction mechanisms for partial and noisy labels, highlight the need for robust learning frameworks to mitigate the compounded effects of and imbalance. Interpretability remains a key hurdle, particularly with deep learning models that dominate high-performance multi-label tasks but operate as black boxes, obscuring the decision rationale behind label predictions. In contrast, tree-based methods offer inherent explainability through hierarchical decision rules, making them preferable for domains requiring transparency, such as healthcare or legal applications, though they often lag in accuracy compared to neural networks. Efforts to bridge this gap, like incorporating tree regularization into deep models, aim to enhance interpretability without sacrificing predictive power, underscoring the trade-off between model complexity and human-understandable reasoning in multi-label settings. Looking ahead, future directions in multi-label classification emphasize privacy-preserving techniques like , which enables collaborative model training across distributed datasets without sharing raw data, addressing concerns in sensitive applications such as . Integration with large language models (LLMs) for zero-shot multi-label classification represents another promising avenue, allowing models to predict labels for unseen categories by leveraging semantic understanding from pre-trained text representations, potentially revolutionizing tasks like automated tagging in dynamic environments. Ethical concerns, including bias amplification in correlated labels—such as demographic attributes in tagging systems—pose additional challenges, where interdependent labels can perpetuate societal prejudices if not mitigated through fair representation learning. These issues necessitate ongoing research into equitable and transparent multi-label systems to ensure responsible deployment.

Resources

Benchmark Datasets

Benchmark datasets play a crucial role in evaluating and comparing multi-label classification algorithms, providing standardized testbeds across diverse domains such as text, bioinformatics, and . These datasets typically feature varying numbers of instances, labels, and degrees of sparsity, reflecting real-world challenges like label correlations and imbalance. Common sources include repositories like the MULAN multi-label dataset collection, which compiles datasets with detailed statistics for . The following table summarizes key characteristics of prominent benchmark datasets, including the number of instances (m), number of labels (q), average number of labels per instance (cardinality), and density (cardinality divided by q):
DatasetDomainmqCardinalityDensityOriginal Source
EnronText (emails)1,702533.3780.064Read et al. (2008) paper
YeastBioinformatics (gene functions)2,417144.2370.303Elisseeff and Weston (2001) paper
SceneComputer vision (landscape images)2,40761.0740.179Boutell et al. (2004) paper
Reuters-21578 (ModApte split)Text (news)10,788901.260.014Lewis (1997) paper
BibtexText (conference papers/tags)7,3951592.4020.015Katakis et al. (2008) paper
AmazonText (product reviews)75,0002,0001.30.00065McAuley et al. (2015) paper
Wiki10-31kText (Wikipedia)17,00030,9264.50.00015Partalas et al. (2015) paper
These datasets exhibit varying sparsity levels; for instance, the dataset is moderately sparse with emails often assigned around 3-4 categories from 53 possible topics, while the dataset is denser, with genes typically linked to over 4 out of 14 functional classes. The Scene dataset, drawn from images, represents a low-label scenario where images are annotated with 6 possible scene elements like mountains or beaches, averaging about 1 label per image. The Reuters-21578 corpus (ModApte split), a large collection of news articles, supports multi-label evaluation with 90 active categories, featuring high sparsity typical of broad text tasks. Similarly, the dataset, derived from social tagging of academic papers, involves 159 tags with very low density, making it suitable for testing algorithms on high-dimensional label spaces. For extreme multi-label settings, datasets like and Wiki10-31k highlight scalability challenges with thousands to tens of thousands of labels and very low densities. These datasets are frequently employed to benchmark problem methods and label correlation models, enabling fair comparisons of algorithmic performance across domains.

Software Implementations

Several open-source libraries and frameworks facilitate the implementation of multi-label classification algorithms across programming languages, enabling practitioners to apply methods like problem transformation and algorithm adaptation to real-world datasets. These tools often integrate with broader ecosystems for preprocessing, evaluation, and deployment. In , the scikit-multilearn library provides comprehensive implementations of problem transformation techniques, including Binary Relevance (BR), Label Powerset (LP), and classifier chains, built on top of for seamless compatibility. itself supports multi-label classification through meta-estimators like MultiOutputClassifier, which wraps binary classifiers, and losses such as binary cross-entropy for extensions. For R users, the utiml package offers a for multi-label , including ensemble methods like classifier chains and random k-labelsets (RAkEL), with support for integrating base learners from other R packages. In Java, the Mulan library implements a wide range of multi-label algorithms from the MULAN , covering , ranking, thresholding, and , with tools for handling multi-label datasets directly. Deep learning frameworks also natively support multi-label tasks. includes the MultiLabelSoftMarginLoss function, which optimizes multi-label one-versus-all predictions using max-entropy principles on multi-hot encoded targets. enables multi-label classification via activations in the output layer combined with binary cross-entropy loss, suitable for neural architectures handling non-exclusive labels. The Transformers library extends this to pre-trained models like , allowing fine-tuning for multi-label text classification by setting the problem type to multi_label_classification and using binary cross-entropy. Many of these libraries integrate with OpenML, an open platform for sharing datasets and experiments, to facilitate reproducible benchmarking of multi-label methods on standardized tasks. For instance, scikit-multilearn's RAkELd ensemble, which combines label powerset with random k-labelsets to capture label dependencies, can be applied to the —a with 2,417 samples and 14 functional classes—by loading the data via scikit-multilearn's utilities, fitting the model with a base classifier like SVM, and evaluating using Hamming loss, achieving competitive performance in label correlation modeling.

References

  1. [1]
    [PDF] A Review on Multi-Label Learning Algorithms - LAMDA
    Multi-label learning studies the problem where each example is represented by a single instance while associated with a set of labels simultaneously.
  2. [2]
  3. [3]
    [PDF] Multi-label Text Classification with a Mixture Model Trained by EM
    Text classification is the problem of assigning a text document into one or more topic categories or classes. In multiclass document classification, as ...
  4. [4]
    Improved Boosting Algorithms Using Confidence-rated Predictions
    We describe several improvements to Freund and Schapire's AdaBoost boosting algorithm, particularly in a setting in which hypotheses may assign confide.
  5. [5]
    Multi‐label learning: a review of the state of the art and ongoing ...
    Nov 22, 2014 · Multi-label learning is quite a recent supervised learning paradigm. Owing to its capabilities to improve performance in problems where a ...
  6. [6]
    [PDF] The utiml Package: Multi-label Classification in R - The R Journal
    ... function f(x, Ai), which, for each class label, outputs a score value in the interval [0.0, 1.0], indicating the relevance, confidence, or probability of.
  7. [7]
    (PDF) Multi-Label Classification: An Overview - ResearchGate
    Aug 7, 2025 · label classification was mainly motivated by the tasks of text categorization and medical diagnosis. Text documents usually belong to more than ...
  8. [8]
    Binary Classification: Spam Filtering - Atmosera
    Jun 23, 2021 · In this post, we'll build a binary-classification model that classifies e-mails as spam or not spam. And we'll use a learning algorithm called Naïve Bayes to ...
  9. [9]
    [PDF] Multi-label Classification Methods: A Comparative Study - IRJET
    Traditional binary and multi-class classifications are two sub-categories of single-label classification. Multi-label classification methods are falling under ...
  10. [10]
    [PDF] Learning Multi-Label Topic Classification of News Articles - CS229
    News article topic classification can enable automatic tagging of articles for online news repositories and the aggregation of news sources by topic (e.g. ...
  11. [11]
    Binary relevance for multi-label learning: an overview
    Mar 23, 2018 · Binary relevance decomposes multi-label learning into independent binary tasks, one per class label, for multi-label examples.Missing: original | Show results with:original
  12. [12]
    [PDF] A Review of Multi-Label Classification Methods - LPIS
    Multi-Label Classification: An Overview. Grigorios Tsoumakas, Ioannis Katakis ... & Vlahavas, I. (2005), 'Protein Classification with Multiple.
  13. [13]
  14. [14]
    Classifier chains for multi-label classification | Machine Learning
    Jun 30, 2011 · A novel classifier chains method that can model label correlations while maintaining acceptable computational complexity.
  15. [15]
    [PDF] Random k-Labelsets for Multi-Label Classification - LPIS
    Therefore, the proposed approach is dubbed RAkEL (RAndom k labELsets). This paper examines the construction of two different types of labelsets: a) disjoint ...
  16. [16]
  17. [17]
    [PDF] Learning Distance Metrics for Multi-Label Classification
    This paper proposes a distance metric learning algorithm for multi-label classification, using Jaccard distance between label vectors, and a loss function ...
  18. [18]
    (PDF) A Review On Multi-Label Learning Algorithms - ResearchGate
    Aug 5, 2025 · This paper aims to provide a timely review on this area with emphasis on state-of-the-art multi-label learning algorithms.
  19. [19]
    Learning rules for multi-label classification: a stacking and a ...
    Apr 29, 2016 · Multi-label classification (MLC) refers to the task of learning a function that maps instances to label subsets or label vectors , where , is a ...
  20. [20]
    Predicting gene function using hierarchical multi-label decision tree ...
    Jan 2, 2010 · Second, a classification tree stores in a leaf the majority class for that leaf; this class will be the tree's prediction for examples ...
  21. [21]
    None
    Summary of each segment:
  22. [22]
    [PDF] CNN-RNN: A Unified Framework for Multi-Label Image Classification
    Microsoft COCO (MS-COCO) dataset [21] is an image recognition, segmentation, and captioning dataset. It con- tains 123 thousand images of 80 objects types with ...
  23. [23]
    AttentionXML: Label Tree-based Attention-Aware Deep Model for ...
    Nov 1, 2018 · Access Paper: View a PDF of the paper titled AttentionXML: Label Tree-based Attention-Aware Deep Model for High-Performance Extreme Multi ...
  24. [24]
    Deep Learning for Extreme Multi-label Text Classification
    Extreme multi-label text classification (XMTC) assigns a document's most relevant subset of class labels from a large collection, where the number of labels ...Missing: definition | Show results with:definition
  25. [25]
    On label dependence and loss minimization in multi-label ...
    Jun 8, 2012 · Thus, by definition, the two labels are conditionally independent, having joint distribution P(Y|x)=P(Y 1|x)×P(Y 2|x) given x. However, ...
  26. [26]
    [PDF] Bayes Optimal Multilabel Classification via Probabilistic Classifier ...
    We propose a probabilistic framework that suggests to look at the problem from the point of view of risk minimization and Bayes opti- mal prediction. Concretely ...
  27. [27]
    Multi-label learning by exploiting label dependency
    In this paper, we propose to use a Bayesian network structure to efficiently encode the conditional dependencies of the labels as well as the feature set.
  28. [28]
    A Tutorial on Multi-label Classification Techniques - ResearchGate
    Aug 9, 2025 · This tutorial presents the most frequently used techniques to deal with these problems in a pedagogical manner, with examples illustrating the main techniques.
  29. [29]
    [PDF] A Study on Threshold Selection for Multi-label Classification
    We introduce the binary method, evaluation measures for multi-label classification, and the basic concept of optimizing evaluation measures in Section 2.Missing: metrics | Show results with:metrics
  30. [30]
    [PDF] AN EMPIRICAL STUDY OF MULTI-LABEL CLASSIFIERS ... - ISMIR
    We make use of the following measures which are commonly used in the multi-label classification literature: Hamming Loss (HL), Average Precision (AP), Coverage ...
  31. [31]
    BoosTexter: A Boosting-based System for Text Categorization
    Schapire, R.E., Singer, Y. BoosTexter: A Boosting-based System for Text Categorization. Machine Learning 39, 135–168 (2000). https://doi.org/10.1023/A ...
  32. [32]
  33. [33]
    Hoeffding adaptive trees for multi-label classification on data streams
    Oct 26, 2024 · This paper proposes a novel approach to multi-label data stream classification called Multi-Label Hoeffding Adaptive Tree (MLHAT).
  34. [34]
    [PDF] Streaming Multi-label Classification
    ADWIN is a change detector and estimator that keeps a variable-length window ... We presented a new experimental framework for multi-label data-stream ...
  35. [35]
    [PDF] Deep Streaming Label Learning
    The proposed framework has three components: 1) streaming label mapping, designed to capture the relationship between new labels and past labels with the ...
  36. [36]
    Review of Extreme Multi-label Classification - arXiv
    May 15, 2025 · Firstly, the large number of labels creates a major bottleneck as it is not possible to have a simple classifier for each label due to memory ...
  37. [37]
    [PDF] Efficient Multi-label Classification with Many Labels
    In this paper, we ad- dress this problem by selecting a small subset of class labels that can approximately span the original label space. This is performed by ...<|control11|><|separator|>
  38. [38]
    Multi-Label Classification with Long-Tailed Distribution and Partial ...
    Apr 20, 2023 · This paper introduces PLT-MLC, a method for multi-label classification with long-tailed distribution and partial labels, using a Correction, ...Missing: noise | Show results with:noise
  39. [39]
    [PDF] A Data Augmentation Strategy against Imbalance and Noise
    BalanceMix is a data augmentation method for multi-label classification that addresses imbalanced and noisy labels by generating minority-augmented instances ...
  40. [40]
    Tree-based explanations of machine learning model predictions
    Increasingly complex learning methods such as boosting, bagging and deep learning have made ML models more accurate, but harder to interpret and explain, ...
  41. [41]
    [PDF] Optimizing for Interpretability in Deep Neural Networks with Tree ...
    Deep models have advanced prediction in many domains, but their lack of interpretability remains a key barrier to the adoption in many real world ...
  42. [42]
    Federated Multi-Label Learning (FMLL): Innovative Method for ...
    Jul 9, 2024 · Furthermore, combining multi-label classification with federated learning (FL) enables distributed and privacy-preserving machine learning ...
  43. [43]
    Large Language Models Do Multi-Label Classification Differently
    May 23, 2025 · We investigate how autoregressive LLMs perform multi-label classification, focusing on subjective tasks, by analyzing the output distributions ...
  44. [44]
    [PDF] ETHICAL CONSIDERATIONS IN MULTILABEL TEXT ...
    We explored different types of biases that creep in during auto-classification: word biases, algorithmic biases, user biases and sensitive categories biases.
  45. [45]
    Multi-Label Classification Dataset Repository
    In this website we provide a huge compilation of multi-label classification datasets, obtained from different sources. For further information, please ...
  46. [46]
    A scikit-learn based module for multi-label et. al. classification - GitHub
    scikit-multilearn is a Python module capable of performing multi-label learning tasks. It is built on-top of various scientific Python packages.
  47. [47]
    scikit-multilearn: Multi-Label Classification in Python — Multi-Label ...
    Scikit-multilearn offers extensive user documentation. Read the user docs, learn from recipes constructed on real data or browse the API reference to find a ...Get started · 1. Exploring Label Relations · Classifier selection · User Guide
  48. [48]
    1.12. Multiclass and multioutput algorithms - Scikit-learn
    Multilabel classification support can be added to any classifier with MultiOutputClassifier . This strategy consists of fitting one classifier per target. This ...
  49. [49]
    Multilabel classification — scikit-learn 1.7.2 documentation
    This example simulates a multi-label document classification problem. The dataset is generated randomly based on the following process.
  50. [50]
    Mulan: A Java library for multi-label learning
    Mulan is an open-source Java library for learning from multi-label datasets. Multi-label datasets consist of training examples of a target function that has ...Getting Mulan · Documentation · Datasets · The Team
  51. [51]
    [PDF] MULAN: A Java Library for Multi-Label Learning
    Abstract. MULAN is a Java library for learning from multi-label data. It offers a variety of classification, rank- ing, thresholding and dimensionality ...
  52. [52]
    Multi-Label Classification with Deep Learning
    Aug 31, 2020 · Multi-label classification involves predicting zero or more class labels. Unlike normal classification tasks where class labels are mutually ...
  53. [53]
    Multi-Label Classification Model From Scratch: Step-by-Step Tutorial
    Jan 8, 2024 · This tutorial will guide you through each step of creating an efficient ML model for multi-label text classification.
  54. [54]
    Python Integrations - Open Machine Learning - OpenML
    Integrating your Python libraries¶. OpenML-Python provides an extension interface to connect other machine learning libraries than scikit-learn to OpenML.Connecting New Machine... · Content Of The Library · Interfacing With...Missing: multi- | Show results with:multi-
  55. [55]
    yeast - OpenML
    The yeast dataset is a multi-label dataset with 2417 genes, 103 predictors, 117 features, 2 classes, and 14 symbolic features.Missing: functions | Show results with:functions
  56. [56]
    scikit-multilearn: A Python library for Multi-Label Classification
    Source code and documentation can be downloaded from http://scikit.ml and also via pip The project is BSD-licensed.