PPL based on trigonometric smoothing functions as activation functions of hidden units
Koji Wada, Makoto Hirai, Shigeyoshi Nakajima, Jiro Okamoto · Osaka City University (Osaka City University) · 2000
Recently Projection Pursuit Learning (PPL) emerged in the Neural Network (NN) literatures in recent years which used smoothing functions as activation functions in hidden units of NN. As this network uses smoothing functions as activation functions for hidden units, adaptability for various problems is expected to be excellent than other networks....