Title :
Acceleration for both Boltzmann machine learning and mean field theory learning
Author :
Hagiwara, Masafumi
Author_Institution :
Psychol. Dept., Stanford Univ., CA, USA
Abstract :
Novel learning algorithms for both the Boltzmann machine (BM) and mean field theory (MFT) are proposed to accelerate learning. The effectiveness of the proposed MFT algorithm is confirmed by computer simulations. It accelerates convergence: for example, the proposed MFT algorithm is more than twice as fast as the conventional MFT algorithm when the learning constant η is 0.1 or less. In addition, it is shown that the proposed algorithm is less sensitive to η. MFT is more biologically plausible and more suitable for VLSI implementation than the other supervised learning paradigms, and it can also be used as a content addressable memory
Keywords :
Boltzmann machines; content-addressable storage; digital simulation; learning (artificial intelligence); Boltzmann machine learning; VLSI implementation; biologically plausible; computer simulations; content addressable memory; convergence; learning constant; mean field theory learning; Acceleration; Artificial neural networks; Computer simulation; Cost function; Energy measurement; Machine learning; Machine learning algorithms; Phase measurement; Psychology; Resonance light scattering;
Conference_Titel :
Neural Networks, 1992. IJCNN., International Joint Conference on
Conference_Location :
Baltimore, MD
Print_ISBN :
0-7803-0559-0
DOI :
10.1109/IJCNN.1992.287107
