Average Information Content Maximization—A New Approach for Fingerprint Hybridization and Reduction

Śmieja, Marek; Warszycki, Dawid

doi:10.1371/journal.pone.0146666

Cited by 15 publications

(21 citation statements)

References 35 publications

(27 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Consequently, we get a third class of compounds with an intermediate activity level such that K i ∈ [10 2 , 10 3 ], which forms a neutral class. Although it is a common practice to ignore this neutral class in the learning process [26], we show that its use allows to explore the chemical space better, see Fig. 1.…”

mentioning

confidence: 81%

SVM with a neutral class

Śmieja

Tabor

Spurek

2017

Pattern Anal Applic

Self Cite

View full text Add to dashboard Cite

a learning framework. For instance, in the semi-supervised learning, the classifier is allowed to use unlabeled data from underlying classes for improving its classification accuracy [5,28]. In universum learning, we might use unlabeled data samples that do not belong to either classes [29,32]. Integrating pre-defined additional information into a learning framework would usually yield improvement in the classification results and obtaining better insight into data.In this paper, we support the above hypothesis and show that neutral instances can be easily handled with a use of Tri-Class SVM model [2]. Our motivation of including neutral class in a training comes from cheminformatics and computer-aided drug design, in which we focus on detecting compounds acting on a particular protein (biological receptor). A compound is considered active if its binding constant K i ∈ [0, +∞) (measured in a laboratory) is lower than a threshold a = 10 2 , while for inactive compounds a binding constant must be greater than b = 10 31 [31]. Consequently, we get a third class of compounds with an intermediate activity level such that K i ∈ [10 2 , 10 3 ], which forms a neutral class. Although it is a common practice to ignore this neutral class in the learning process [26], we show that its use allows to explore the chemical space better, see Fig. 1.Tri-Class SVM [2] is a generalization of classical SVM [8], which builds a single learning model for three-class problems and avoids pairwise coupling strategy. To use instances of neutral class in the learning process, we develop its two parameterizations: SVM {0} and SVM [−1,1] . In analogy to classical SVM, we look for such a hyperplane which maximizes the margin between positive and negative examples Abstract In many real binary classification problems, in addition to the presence of positive and negative classes, we are also given the examples of third neutral class, i.e., the examples with uncertain or intermediate state between positive and negative. Although it is a common practice to ignore the neutral class in a learning process, its appropriate use can lead to the improvement in classification accuracy. In this paper, to include neutral examples in a training stage, we adapt two variants of Tri-Class SVM (proposed by Angulo et al. in Neural Process Lett 23(1):89-101, 2006), the method designed to solve three-class problems with a use of single learning model. In analogy to classical SVM, we look for such a hyperplane, which maximizes the margin between positive and negative instances and which is localized as close to the neutral class as possible. In addition to original Angulo's paper, we give a new interpretation of the model and show that it can be easily implemented in the primal. Our experiments demonstrate that considered methods obtain better results in binary classification problems than classical SVM and semi-supervised SVM.

show abstract

mentioning

confidence: 81%

SVM with a neutral class

Śmieja

Tabor

Spurek

2017

Pattern Anal Applic

Self Cite

View full text Add to dashboard Cite

show abstract

“…To resolve the aforementioned difficulties with application of high resolution fingerprints, the AIC-MAX algorithm [19] was recently introduced to select features with the highest discriminatory potential in virtual screening-like experiments. AIC-MAX uses mutual information normalized by the Shannon entropy to rank a group of features with respect to their significance measured by activity label .where is a binary sequence (fingerprint of length N ) and , and denote the probabilities that , } and , , respectively.…”

Section: Methodsmentioning

confidence: 99%

Practical application of the Average Information Content Maximization (AIC-MAX) algorithm: selection of the most important structural features for serotonin receptor ligands

2017

Self Cite

View full text Add to dashboard Cite

The Average Information Content Maximization algorithm (AIC-MAX) based on mutual information maximization was recently introduced to select the most discriminatory features. Here, this methodology was applied to select the most significant bits from the Klekota-Roth fingerprint for serotonin receptors ligands as well as to select the most important features for distinguishing ligands with activity for one receptor versus another. The interpretation of selected bits and machine-learning experiments performed using the reduced interpretations outperformed the raw fingerprints and indicated the most important structural features of the analyzed ligands in terms of activity and selectivity. Moreover, the AIC-MAX methodology applied here for serotonin receptor ligands can also be applied to other target classes.Electronic supplementary materialThe online version of this article (doi:10.1007/s11030-017-9729-8) contains supplementary material, which is available to authorized users.

show abstract

“…Step 4: Update the velocity by using Equation (8) and the binary position by using Equations (10) and (11).…”

Section: Binary Particle Swarm Optimizationmentioning

confidence: 99%

“…In [7,8], the Fisher criterion is utilized to evaluate features comprehensively, considering both the intra-class distance and the inter-class distance. Information measure-based feature selection methods are utilized to select important features in [9,10]. Other methods, such as the relief method [11] and the rough set method [12], are also adopted for TSSP feature selection.…”

Section: Introductionmentioning

confidence: 99%

A Two-Stage Feature Selection Method for Power System Transient Stability Status Prediction

Chen¹,

Han

Fan³

et al. 2019

Energies

View full text Add to dashboard Cite

Transient stability status prediction (TSSP) plays an important role in situational awareness of power system stability. One of the main challenges of TSSP is the high-dimensional input feature analysis. In this paper, a novel two-stage feature selection method is proposed to handle this problem. In the first stage, the relevance between features and classes is measured by normalized mutual information (NMI), and the features are ranked based on the NMI values. Then, a predefined number of top-ranked features are selected to form the strongly relevant feature subset, and the remaining features are described as the weakly relevant feature subset, which can be utilized as the prior knowledge for the next stage. In the second stage, the binary particle swarm optimization is adopted as the search algorithm for feature selection, and a new particle encoding method that considers both population diversity and prior knowledge is presented. In addition, taking the imbalanced characteristics of TSSP into consideration, an improved fitness function for TSSP feature selection is proposed. The effectiveness of the proposed method is corroborated on the Northeast Power Coordinating Council (NPCC) 140-bus system.

show abstract

Average Information Content Maximization—A New Approach for Fingerprint Hybridization and Reduction

Cited by 15 publications

References 35 publications

SVM with a neutral class

SVM with a neutral class

Practical application of the Average Information Content Maximization (AIC-MAX) algorithm: selection of the most important structural features for serotonin receptor ligands

A Two-Stage Feature Selection Method for Power System Transient Stability Status Prediction

Contact Info

Product

Resources

About