Self-Learning Optical Neural Network

Abstract

One of the features in neural computing must be the adaptability to changeable environment and to recognize unknown objects. In general, there are two types of learning processes that are used in the human brain; supervised and unsupervised learnings [1]. In a supervised learning process, the artificial neural network has to be taught when to learn and when to process the information. Nevertheless, if an unknown object is presented to the artificial neural network during the processing, the network may provide an error output result. On the other hand, for unsupervised learning (also called self-learning), the students are learning by themselves, in which based on simple learning rules and their past experiences. Kohonon's model is one of the simplest self-organizing algorithms[1], which is capable of performing statistical pattern recognition and classification, and it can be modified for optical neural network implementation. A compact optical neural network of 64 neurons using liquid crystal televisions is used for unsupervised learning process[2].