Distinct Chains for Different Instances: An Effective Strategy for Multi-label Classifier Chains

Silva, Pablo Nascimento da; Gonçalves, Eduardo Corrêa; Plastino, Alexandre; Freitas, Alex A.

doi:10.1007/978-3-662-44851-9_29

Cited by 15 publications

(16 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A common characteristic of these proposals is that they try to eliminate a key drawback in the original method: the fact that the label ordering is decided at random. In a recent study [3], an exhaustive experiment that compared the predictive performance of CC models built with thousands of distinct label orderings revealed that the differences in accuracy tend to be very large. The same study also revealed that, for many datasets, the original CC method (trained with a random order) might present a performance inferior to the BR method even in problems where most labels are correlated with each other.…”

Section: Related Workmentioning

confidence: 98%

“…The authors of the original CC model proposed the combination of random orders via an ensemble of classifier chains (ECC) in order to mitigate the effect of poorly ordered chains [22]. Differently, the approaches proposed in [3,11,16,21] employ heuristic methods. In [16], the LSOP is addressed by performing a beam search over a tree in which every distinct path represents a different label permutation.…”

Section: Related Workmentioning

confidence: 99%

“…Crossover works by transferring sub-chains of random length between two individuals whilst mutation swaps pairs of labels of an individual. Finally, [3] proposes a lazy approach capable of searching for a distinct and more effective label sequence to each new instance t. In this strategy, for each instance in the training dataset, an effective label sequence is previously identified. At classification time, the label sequence for the new instance t is chosen based on the sequences associated with the instances in the training dataset that are more similar to t.…”

Section: Related Workmentioning

confidence: 99%

“…A number of variations of the CC method have recently been proposed [3,11,16,21,24]. A common characteristic of these proposals is that they try to eliminate a key drawback in the original method: the fact that the label ordering is decided at random.…”

Section: Related Workmentioning

confidence: 99%

“…First, it decides the label sequence at random. As demonstrated in [3], the use of distinct chain orderings can lead to large differences in the predictive accuracy of the model. Second, CC forces all labels to be present in the chain, despite the fact that some of them might carry redundant and irrelevant information.…”

Section: Introductionmentioning

confidence: 98%

See 4 more Smart Citations

Simpler is Better

Gonçalves

Plastino

Freitas

2015

Proceedings of the 2015 Annual Conference on Genetic and Evolutionary Computation

Self Cite

View full text Add to dashboard Cite

Multi-label classification (MLC) is the task of assigning multiple class labels to an object based on the features that describe the object. One of the most effective MLC methods is known as Classifier Chains (CC). This approach consists in training q binary classifiers linked in a chain, {y1 → y2 → ... → yq}, with each responsible for classifying a specific label in {l1, l2, ..., lq}. The chaining mechanism allows each individual classifier to incorporate the predictions of the previous ones as additional information at classification time. Thus, possible correlations among labels can be automatically exploited. Nevertheless, CC suffers from two important drawbacks: (i) the label ordering is decided at random, although it usually has a strong effect on predictive accuracy; (ii) all labels are inserted into the chain, although some of them might carry irrelevant information to discriminate the others. In this paper we tackle both problems at once, by proposing a novel genetic algorithm capable of searching for a single optimized label ordering, while at the same time taking into consideration the utilization of partial chains. Experiments on benchmark datasets demonstrate that our approach is able to produce models that are both simpler and more accurate.

show abstract

Section: Related Workmentioning

confidence: 98%

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 98%

See 3 more Smart Citations

Simpler is Better

Gonçalves

Plastino

Freitas

2015

Proceedings of the 2015 Annual Conference on Genetic and Evolutionary Computation

Self Cite

View full text Add to dashboard Cite

show abstract

Dynamic Classifier Chain with Random Decision Trees

Kulessa

Mencía

2018

Discovery Science

View full text Add to dashboard Cite

Dynamic Classifier Chains for Multi-label Learning

Trajdos

Kurzyński

2019

Lecture Notes in Computer Science

View full text Add to dashboard Cite

In this paper, we deal with the task of building a dynamic ensemble of chain classifiers for multi-label classification. To do so, we proposed two concepts of classifier chains algorithms that are able to change label order of the chain without rebuilding the entire model. Such modes allows anticipating the instance-specific chain order without a significant increase in computational burden. The proposed chain models are built using the Naive Bayes classifier and nearest neighbour approach as a base single-label classifiers. To take the benefits of the proposed algorithms, we developed a simple heuristic that allows the system to find relatively good label order. The heuristic sort labels according to the label-specific classification quality gained during the validation phase. The heuristic tries to minimise the phenomenon of error propagation in the chain. The experimental results showed that the proposed model based on Naive Bayes classifier the above-mentioned heuristic is an efficient tool for building dynamic chain classifiers.

show abstract

Distinct Chains for Different Instances: An Effective Strategy for Multi-label Classifier Chains

Cited by 15 publications

References 20 publications

Simpler is Better

Simpler is Better

Dynamic Classifier Chain with Random Decision Trees

Dynamic Classifier Chains for Multi-label Learning

Contact Info

Product

Resources

About