DocumentCode :
226615
Title :
Regularization-based learning of the Choquet integral
Author :
Anderson, Derek T. ; Price, Stanton R. ; Havens, Timothy C.
Author_Institution :
Dept. of Electr. & Comput. Eng., Mississippi State Univ., Starkville, MS, USA
fYear :
2014
fDate :
6-11 July 2014
Firstpage :
2519
Lastpage :
2526
Abstract :
A number of data-driven fuzzy measure (FM) learning techniques have been put forth for the fuzzy integral (FI). Examples include quadratic programming, Gibbs sampling, gradient descent, reward and punishment and evolutionary optimization. However, most approaches focus solely on the minimization of the sum of squared error (SSE). Limited attention has been placed on characterizing and subsequently minimizing model (i.e., FM) complexity. Furthermore, the vast majority of learning techniques are highly susceptible to over-fitting and noise. Herein, we explore a regularization approach to learning the FM for the Choquet FI. We investigate the mathematical motivation for such an approach, its applicability and impact on different types of FMs, and its desirable properties for quadratic programming (QP) based optimization. We show that L regularization has a distinct meaning for measure learning and aggregation operators. Experiments are performed and validated with respect to the Shapley index. Specifically, we show that it is possible to reduce the effect of overfitting, we can identify higher quality measures and, if desired, force the learning of fewer numbers of sources.
Keywords :
fuzzy set theory; learning (artificial intelligence); quadratic programming; Choquet FI; Choquet integral; FM learning techniques; L1 regularization; QP based optimization; SSE; Shapley index; aggregation operators; data-driven fuzzy measure learning techniques; fuzzy integral; overfitting reduction; quadratic programming based optimization; regularization-based learning; sum of squared error minimization; Complexity theory; Equations; Frequency modulation; Indexes; Mathematical model; Noise; Vectors;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Fuzzy Systems (FUZZ-IEEE), 2014 IEEE International Conference on
Conference_Location :
Beijing
Print_ISBN :
978-1-4799-2073-0
Type :
conf
DOI :
10.1109/FUZZ-IEEE.2014.6891630
Filename :
6891630
Link To Document :
بازگشت