Plenary Lecture
The Use of Meta-Optimization for Parameter Selection in Machine Learning
Professor Filippo Neri
Dept. of Electrical Engineering and Computer Science
University of Naples Federico II
Italy
E-mail: filipponeri@yahoo.com
Abstract: The process of identifying the optimal parameters for an optimization algorithm or a machine learning one is a computationally expensive one and it usually requires the search of a large, possibly infinite, space of candidate parameter sets. This process does not have any guarantee of optimality. Various attempts have been made to automate this process. I will describe my current research in the field by describing a methodological approach of using a simple genetic algorithm to approximate the optimal parameter setting for machine learning system on given datasets.
Brief Biography of the Speaker: Prof. Filippo Neri is currently with the Dept. of Electrical Engineering and Computer Science at University of Naples Federico II, Italy. Prof. Filippo Neri is currently Editor in Chief of WSEAS Transactions on Systems. Prof. Filippo Neri has wide experience in the area of artificial intelligence, machine learning, and software agent simulation. He had the opportunity to work both in academic and industrial environments including Ericsson's and Unlever's R&D centers and across three countries in the European Union (Italy, Ireland and UK). He has studied and visited at several important academic institutions including Carnegie Mellon University, Imperial College London, University of Milano, University of Torino, University of Malta. He is a Marie Curie Fellow and a ADI associate, the Italian PhD association. Finally he has served in the program committees and as reviewer at several international conferences.