A New Hybrid Method for Feature Selection using Ant Colony Optimization and Firefly Algorithm
Publish place: International Conference on Engineering, Art and Environment
Publish Year: 1393
نوع سند: مقاله کنفرانسی
زبان: English
View: 776
This Paper With 13 Page And PDF Format Ready To Download
- Certificate
- من نویسنده این مقاله هستم
استخراج به نرم افزارهای پژوهشی:
شناسه ملی سند علمی:
CEAE01_221
تاریخ نمایه سازی: 14 مرداد 1394
Abstract:
In high-dimensional feature space, unrelated features decrease the accuracy of classification and increase the computational complexity. The Selection of feature provides the relevant and useful information and improves the efficiency. The Selection of feature can be seen as an optimization problem because the selection of appropriate subset of features is very important. This paper presents a novel feature selection algorithm based on combination of Ant Colony optimization and firefly algorithm, called FFACO to optimize the selection of features. The Ant Colony algorithm is a famous meta-heuristic search algorithm used in solving combinatorial optimization problems. Firefly algorithm is an evolutionary model based on collective intelligence algorithms and derived from nature. This algorithm is mainly used in solving optimization problem. Eight UCI datasets and three classifier learning algorithm have been used for evaluating the proposed algorithm. Experimental results show that proposed algorithm (FFACO), increases classification accuracy, by selecting the least number of features, in most instances
Keywords:
Authors
Esmat Zandi
Department of Computer Engineering, Isfahan (Khorasgan) Branch, Islamic Azad University, Isfahan, Iran.
Mehdi Sadeghzadeh
Department of Computer Engineering, college of electronic and computer Mahshahr Branch, Islamic Azad University, Mahshahr, Iran.
مراجع و منابع این Paper:
لیست زیر مراجع و منابع استفاده شده در این Paper را نمایش می دهد. این مراجع به صورت کاملا ماشینی و بر اساس هوش مصنوعی استخراج شده اند و لذا ممکن است دارای اشکالاتی باشند که به مرور زمان دقت استخراج این محتوا افزایش می یابد. مراجعی که مقالات مربوط به آنها در سیویلیکا نمایه شده و پیدا شده اند، به خود Paper لینک شده اند :