META-DES.Oracle: Meta-learning and feature selection for dynamic ensemble selection

Abstract

Dynamic ensemble selection (DES) techniques work by estimating the competence level of each classifier from a pool of classifiers, and selecting only the most competent ones for the classification of a specific test sample. The key issue in DES is defining a suitable criterion for calculating the classifiers' competence. There are several criteria available to measure the level of competence of base classifiers, such as local accuracy estimates and ranking. However, using only one criterion may lead to a poor estimation of the classifier's competence. In order to deal with this issue, we have proposed a novel dynamic ensemble selection framework using meta-learning, called META-DES. A meta-classifier is trained, based on the meta-features extracted from the training data, to estimate the level of competence of a classifier for the classification of a given query sample. An important aspect of the META-DES framework is that multiple criteria can be embedded in the system encoded as different sets of meta-features. However, some DES criteria are not suitable for every classification problem. For instance, local accuracy estimates may produce poor results when there is a high degree of overlap between the classes. Moreover, a higher classification accuracy can be obtained if the performance of the meta-classifier is optimized for the corresponding data. In this paper, we propose a novel version of the META-DES framework based on the formal definition of the Oracle, called META-DES.Oracle. The Oracle is an abstract method that represents an ideal classifier selection scheme. A meta-feature selection scheme using an overfitting cautious Binary Particle Swarm Optimization (BPSO) is proposed for improving the performance of the meta-classifier. The difference between the outputs obtained by the meta-classifier and those presented by the Oracle is minimized. Thus, the meta-classifier is expected to obtain results that are similar to the Oracle. Experiments carried out using 30 classifi-1 arXiv:1811.00217v1 [cs.LG] 1 Nov 2018 cation problems demonstrate that the optimization procedure based on the Oracle definition leads to a significant improvement in classification accuracy when compared to previous versions of the META-DES framework and other state-of-the-art DES techniques.

Dynamic ensemble selection VS K-NN: Why and when dynamic selection obtains higher classification performance?

“…For DCS, the following techniques were evaluated: Local Class Accuracy (LCA) [7], Overall Local Accuracy (OLA) [7], Modified Local Accuracy (MLA) [8], Modified Classifier Ranking (RANK) [10], [7], Multiple Classifier Behavior (MCB) [20], A Priori [18], [19], A Posteriori [18], [19] and the Dynamic Selection on Complexity (DSOC). For dynamic ensemble selection, the following techniques were considered: K-Nearest Oracles Eliminate (KNORA-E) [4], K-Nearest Oracles Union (KNORA-U) [25], Randomized Reference Classifier (DES-RRC) [28], K-Nearest Output Profiles (KNOP) [25], [29], Dynamic Ensemble Selection Performance (DES-P) [11], Dynamic Ensemble Selection Kullback-Leibler (DES-KL) [11], DES Clustering [9], DES-KNN [9], Meta Learning for Dynamic Selection (META-DES) [5] and META-DES.Oracle [30].…”

FIRE-DES++: Enhanced online pruning of base classifiers for dynamic ensemble selection

“…We used 8 dynamic classifier selection techniques from the literature. (Table 1) In addition, we compare the proposed FIRE-DES++ with the three dynamic ensemble selection frameworks that achieved the best classification performance in [1]: Randomized Reference Classifier (RRC) [12], META-DES [13], and META-DES.Oracle [14]. They are briefly described below:…”

“…The META-DES framework has two additional hyper-parameters: The number of samples selected using output profiles K p and the sample selection threshold h c . The values of the hyper-parameters K p and h c for the META-DES framework were set to 5 and 80% according to the results presented in [13,14].…”

“…We evaluated FIRE-DES++ on 8 dynamic selection techniques: Overall Local Accuracy (OLA) [7], Local Class Accuracy(LCA) [7], A Priori selection [8], A Posteriori selection [8], Multiple Classifier Behavior (MCB) [9], Dynamic Selection KNN [10] and the K-Nearest Oracles Union (KNU) and Eliminate (KNE) [11]. We also compared FIRE-DES++ with the better performing dynamic selection technique according to a recent survey [1]: Randomized Reference Classifier (RRC) [12], META-DES [13], and META-DES.Oracle [14] as well as several static ensemble approaches. This paper is organized as follows: Section 2 presents the problem statement, Section 3 presents the proposed framework, Section 4 presents the experimental study, and Section 5 concludes the paper.…”

Cross-project defect prediction with meta-Learning

“…In this context, the meta-learning task is commonly designed to produce a rank of possible solutions [22]. The recommended order can be useful either to guide the decision task [51] or to compose ensemble solutions [52].…”