S-Shaped vs. V-Shaped Transfer Functions for Ant Lion Optimization Algorithm in Feature Selection Problem

These days, a massive quantity of data is produced online and is incorporated into a variety of datasets in the form of features, however there are lot of features in these datasets that may not be relevant to the problem. In this perspective, feature selection aids to improve the classification accuracy with lesser number of features, which can be well thought-out as an optimization problem. In this paper, Sine Cosine Algorithm (SCA) hybridized with Ant Lion Optimizer (ALO) to form a hybrid Sine Cosine Ant Lion Optimizer (SCALO) is proposed. The proposed algorithm is mapped to its binary versions by using the concept of transfer functions, with the objective to eliminate the inappropriate features and to enhance the accuracy of the classification algorithm (or in any case remains the same). For the purpose of experimentation, this research considers 18 diverse datasets and moreover, the performance of the binary versions of SCALO is compared with some of the latest metaheuristic algorithms, on the basis of various criterions. It can be observed that the binary versions of SCALO perform better than the other algorithms on various evaluation criterions for solving feature selection problem.

Download Full-text

Development of an Enhanced Ant Lion Optimization Algorithm and its Application in Antenna Array Synthesis

Applied Soft Computing ◽

10.1016/j.asoc.2017.05.007 ◽

2017 ◽

Vol 59 ◽

pp. 153-173 ◽

Cited By ~ 18

Author(s):

K.R. Subhashini ◽

J.K. Satapathy

Keyword(s):

Antenna Array ◽

Optimization Algorithm ◽

Ant Lion Optimization ◽

Ant Lion ◽

Lion Optimization Algorithm ◽

Antenna Array Synthesis

Download Full-text

U-curve: A branch-and-bound optimization algorithm for U-shaped cost functions on Boolean lattices applied to the feature selection problem

Pattern Recognition ◽

10.1016/j.patcog.2009.08.018 ◽

2010 ◽

Vol 43 (3) ◽

pp. 557-568 ◽

Cited By ~ 14

Author(s):

Marcelo Ris ◽

Junior Barrera ◽

David C. Martins

Keyword(s):

Feature Selection ◽

Branch And Bound ◽

Optimization Algorithm ◽

Selection Problem ◽

Cost Functions ◽

Feature Selection Problem ◽

Boolean Lattices ◽

Bound Optimization

Download Full-text

A comparative analysis of the queuing search algorithm, the sine-cosine algorithm, the ant lion algorithm to determine the optimal weight design problem of a spur gear drive system

Materials Testing ◽

10.1515/mt-2020-0075 ◽

2021 ◽

Vol 63 (5) ◽

pp. 442-447

Author(s):

Hammoudi Abderazek ◽

Ferhat Hamza ◽

Ali Riza Yildiz ◽

Liang Gao ◽

Sadiq M. Sait

Keyword(s):

Optimization Algorithm ◽

Search Algorithm ◽

Spur Gear ◽

Ant Lion Optimization ◽

Lion Algorithm ◽

Weight Design ◽

Sine Cosine Algorithm ◽

Ant Lion ◽

Teaching Learning ◽

Lion Optimization Algorithm

Abstract Metaheuristic optimization algorithms have gained relevance and have effectively been investigated for solving complex real design problems in diverse fields of science and engineering. In this paper, a recent meta-heuristic approach inspired by human social concepts, namely the queuing search algorithm (QSA), is implemented for the first time to optimize the main parameters of the spur gear, in particular, to minimize the weight of a single-stage spur gear. The effectiveness of the algorithm introduced is examined in two steps. First, the algorithm used is compared with descriptions in previous studies and indicates that the final results obtained by QSA lead to a reduction in gear weight by 7.5 %. Furthermore, the outcomes obtained are compared with those for the other five algorithms. The results reveal that the QSA outperforms the techniques with which it is compared such as the sine-cosine optimization algorithm, the ant lion optimization algorithm, the interior search algorithm, the teaching-learning-based algorithm, and the jaya algorithm in terms of robustness, success rate, and convergence capability.

Download Full-text