Optimized bio-inspired kernels with twin support vector machine using low identity sequences to solve imbalance multiclass classification
| Autor: | Rd. Rohmat Saedudin, Baraa Wasfi Salim, Shahreen Kasim, Rohayanti Hassan, S.K. Guramand, Rohaizan Ramlan |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
Environmental Engineering
genetic structures business.industry Computer science Health Toxicology and Mutagenesis Spectrum (functional analysis) Process (computing) food and beverages Pattern recognition Function (mathematics) Toxicology Protein evolution Support vector machine Multiclass classification Distribution (mathematics) Identity (object-oriented programming) Artificial intelligence business |
| Zdroj: | Journal of Environmental Biology. 40:563-576 |
| ISSN: | 2394-0379 0254-8704 |
| Popis: | The function of enzymes is performed differently depending on their bio-chemical mechanisms and important to the prediction of protein structure and function. In order to overcome the weaknesses of imbalance data distribution in subclasses prediction we proposed Bio-Twin Support Vector Machine (Bio–TWSVM). The TWSVM approach as also allow for kernel optimization where in this study we have introduced the bio-inspired kernels such as the Fisher, spectrum and mismatch kernels which at the same time incorporate the biological information regarding the protein evolution in the classification process. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|