Tuning Parameters in Deep Belief Networks for Time Series Prediction through Harmony Search

There have been several researches of applying Deep Belief Networks (DBNs) to predict time series data. Most these works pointed out that DBNs can bring better prediction accuracy than traditional Artificial Neural Networks. However, one the main shortcomings using in concerns with proper selection their parameters. In this paper, we investigate use Harmony Search algorithm for determining parameters DBN forecasting series. Experimental results on synthetic and real world datasets revealed selected by performs Particle Swarm Optimization (PSO) or random method most tested datasets.