Improving obsolescence detection accuracy using recurrent neural networks

Abstract

Forecasting a product's obsolescence depends on a multitude of factors which can be both technical and non-technical aspects of the product under study. The predictions are usually an approximate of the obsolescence and might not reflect the true nature of the product. Thus, researchers from various fields including market research, technology, public perception and others unite together in order to device a model which can be used for efficient obsolescence detection of products. In this paper, we propose an algorithm for effective obsolescence detection with the help of integrated datasets and a recurrent neural network (RNN). The RNN is used so that the effectiveness of prediction can be improved, and it is found that RNN is better when compared with other standard prediction classifiers.