Proceedings of the BRNS sponsored theme meeting on artificial neural network

Neural network (NN) has emerged as a very powerful calculation tool. It doesn't follow the traditional algorithmic approach for solving a problem. It has to be trained first for solving a specific problem - like human beings. NNs are generally used for solving complex problems including ones' in nuclear industry, which are very difficult to express in algorithmic manner. They learn from examples from the given problem domain and become expert. Later, if an input set is given to the trained NN, which was never used during training, the NN can correctly predict the corresponding output. NNs have been used for solving many complex problems in nuclear industry, speech recognition, pattern matching, classification, optimisation, time series analysis, control, noise filtering etc. Papers relevant INIS are indexed separately