Publication details


G. Ferrari-Trecate and M. Muselli. A new learning method for piecewise linear regression. Artificial Neural Networks - ICANN 2002: International Conference, Madrid, Spain, August 28-30, 2002. Proceedings. Lecture Notes in Computer Science , 2623:444--449, 2002. Madrid, Spain.


A new connectionist model for the solution of piecewise lin-ear regression problems is introduced; it is able to reconstruct both con-tinuous and non continuous real valued mappings starting from a finiteset of possibly noisy samples. The approximating function can assume adifferent linear behavior in each region of an unknown polyhedral parti-tion of the input domain.The proposed learning technique combines local estimation, clustering inweight space, multicategory classification and linear regression in orderto achieve the desired result. Through this approach piecewise affinesolutions for general nonlinear regression problems can also be found.