Title :
Application of entropic value-at-risk in machine learning with corrupted input data
Author :
Ahmadi-Javid, Amir
Author_Institution :
Dept. of Ind. Eng., Amirkabir Univ. of Technol., Tehran, Iran
Abstract :
The entropic value-at-risk (EVaR) is a coherent risk measure that is efficiently computable for the sum of independent random variables. This paper shows how this risk measure can be used in machine learning when uncertainty affects the input data. For this purpose, we consider here a support vector machine with corrupted input data.
Keywords :
learning (artificial intelligence); random processes; support vector machines; EVaR; corrupted input data; entropic value-at-risk; machine learning; random variable; risk measure; support vector machine; Linear programming; Machine learning; Optimization; Random variables; Reactive power; Support vector machines; Uncertainty;
Conference_Titel :
Information Science, Signal Processing and their Applications (ISSPA), 2012 11th International Conference on
Conference_Location :
Montreal, QC
Print_ISBN :
978-1-4673-0381-1
Electronic_ISBN :
978-1-4673-0380-4
DOI :
10.1109/ISSPA.2012.6310455
