Title :
The α-EM learning and its cookbook: from mixture-of-expert neural networks to movie random field
Author :
Matsuyama, Yasuo ; Ikeda, Takayuki ; Tanaka, Tomoaki ; Furukawa, Satoshi ; Takeda, Naoki ; Niimoto, Takeshi
Author_Institution :
Dept. of Electr., Electron. & Comput. Eng., Waseda Univ., Tokyo, Japan
Abstract :
The α-EM algorithm is a proper extension of the traditional log-EM algorithm. This new algorithm is based on the α-logarithm, while the traditional one uses the logarithm. The case of α=-1 corresponds to the log-EM algorithm. Since the speed of the α-EM algorithm was reported for learning problems, this paper shows that closed-form E-steps can be obtained for a wide class of problems. There is a set of common techniques. That is, a cookbooks for the α-EM algorithm is presented. The recipes include unsupervised neural networks, supervised neural networks for various gating, hidden Markov models and Markov random fields for moving object segmentation. Reasoning for the speedup is also given
Keywords :
learning (artificial intelligence); maximum likelihood estimation; neural nets; α-EM learning; Markov random fields; closed-form E-steps; expectation-maximization algorithms; gating; hidden Markov models; log-EM algorithm; mixture-of-expert neural networks; movie random field; moving object segmentation; supervised neural networks; unsupervised neural networks; Clustering algorithms; Equations; Motion pictures; Neural networks; Supervised learning;
Conference_Titel :
Neural Networks, 1999. IJCNN '99. International Joint Conference on
Conference_Location :
Washington, DC
Print_ISBN :
0-7803-5529-6
DOI :
10.1109/IJCNN.1999.831162
