Monte Carlo simulation based algorithm design for automatic learning machine performance analysis

Author

Tigani, Smail ; Ouzzif, Mouhamed ; Hasbi, Abderrahim

Author_Institution

Comput. Sci. Dept., Nat. High Sch. of Electr. & Mech., Casablanca, Morocco

fYear

2014

fDate

28-30 May 2014

Firstpage

19

Lastpage

21

Abstract

This paper proposes advanced key performance indicators dedicated to learning machines and auto-adaptive systems performance analysis. This work introduces an algorithm implementing designed key performance indicators for automatic learning capacity checking. The algorithm simulates the supervised environment to stimulate the tested auto-adaptive machine and then study its adaption capacity based on indicators statistically designed.

Keywords

Monte Carlo methods; learning (artificial intelligence); Monte Carlo simulation based algorithm design; advanced key performance indicators; autoadaptive machine; autoadaptive system performance analysis; automatic learning capacity checking; automatic learning machine performance analysis; Algorithm design and analysis; Linear regression; Mathematical model; Monte Carlo methods; Prediction algorithms; Predictive models; Vectors; Automatic Learning; Key Performance Indicator; Linear Regression; Monte Carlo Simulation; Prediction Algorithms;

fLanguage

English

Publisher

ieee

Conference_Titel

Next Generation Networks and Services (NGNS), 2014 Fifth International Conference on

Conference_Location

Casablanca

Print_ISBN

978-1-4799-6608-0

Type

conf

DOI

10.1109/NGNS.2014.6990221

Filename

6990221