Feature selection approach based on particle swarm optimization algorithm in virtual screening process / Abderrahim Nouioua  

Public ISBD Titre : Feature selection approach based on particle swarm optimization algorithm in virtual screening process Type de document : texte imprimé Auteurs : Abderrahim Nouioua, Auteur ; Fouaz Berrhail, Directeur de thèse Année de publication : 2022 Importance : 1 vol (807 f .) Format : 29cm Langues : Français (fre) Catégories : Thèses & Mémoires:Informatique Mots-clés : Ligand-Based Virtual Screening Similarity Searching Index. décimale : 004 Informatique Résumé : In the last years, similarity searching has gained wide popularity as a method for performing ligand-based virtual screening (LBVS). This screening technique functions by making a comparison of the target compound’s features with that of each compound in the database of compounds. The Tanimoto coefficient is employed to determine the similarity between a database structure and a target structure. This coefficient is considered currently the most widely used coefficient in chemical information systems of fingerprint-based similarity.Different techniques can be employed to improve the performances of the virtual screening process. In this work, our main objective is to propose a feature selection approach based on particle swarm optimization algorithm to improve to performance of the screening process in chemical database. Our contribution consist of identify the most important and relevant features of chemical compounds and to minimize their number in their representations. Our proposed method allow at first; the reduction in features space, the elimination of redundancy and the decrease in training run-time, and secondly to boost performances and effectiveness of the screening process. After performing several experiments our approach was able to outperform the standard Tanimoto method, and we were able to increase the effectiveness of the similarity searching methods, by eliminate the irrelevant and redundant features from the dataset. The achieved results given for all data sets at the top 1% and 5% are proof of the reliability of our proposals. Thus, the proposed approach performs better compared to the Tanimoto method. Côte titre : MAI/0572 En ligne : https://drive.google.com/file/d/174OrvQVu7IhVKifrvfjpwUtYToVSpFEd/view?usp=share [...] Format de la ressource électronique : pdf Feature selection approach based on particle swarm optimization algorithm in virtual screening process [texte imprimé] / Abderrahim Nouioua, Auteur ; Fouaz Berrhail, Directeur de thèse . - 2022 . - 1 vol (807 f .) ; 29cm. Langues : Français (fre) Catégories : Thèses & Mémoires:Informatique Mots-clés : Ligand-Based Virtual Screening Similarity Searching Index. décimale : 004 Informatique Résumé : In the last years, similarity searching has gained wide popularity as a method for performing ligand-based virtual screening (LBVS). This screening technique functions by making a comparison of the target compound’s features with that of each compound in the database of compounds. The Tanimoto coefficient is employed to determine the similarity between a database structure and a target structure. This coefficient is considered currently the most widely used coefficient in chemical information systems of fingerprint-based similarity.Different techniques can be employed to improve the performances of the virtual screening process. In this work, our main objective is to propose a feature selection approach based on particle swarm optimization algorithm to improve to performance of the screening process in chemical database. Our contribution consist of identify the most important and relevant features of chemical compounds and to minimize their number in their representations. Our proposed method allow at first; the reduction in features space, the elimination of redundancy and the decrease in training run-time, and secondly to boost performances and effectiveness of the screening process. After performing several experiments our approach was able to outperform the standard Tanimoto method, and we were able to increase the effectiveness of the similarity searching methods, by eliminate the irrelevant and redundant features from the dataset. The achieved results given for all data sets at the top 1% and 5% are proof of the reliability of our proposals. Thus, the proposed approach performs better compared to the Tanimoto method. Côte titre : MAI/0572 En ligne : https://drive.google.com/file/d/174OrvQVu7IhVKifrvfjpwUtYToVSpFEd/view?usp=share [...] Format de la ressource électronique : pdf

Exemplaires (1)

Code-barres	Cote	Support	Localisation	Section	Disponibilité
MAI/0572	MAI/0572	Mémoire	Bibliothéque des sciences	Français	Disponible Disponible

---

1 (1 - 1 / 1)