PROJECT TITLE :

Toward Automatic Time-Series Forecasting Using Neural Networks

ABSTRACT:

Over the past few decades, application of Artificial Neural Networks (ANN) to time-series forecasting (TSF) has been growing rapidly due to several unique features of ANN models. However, up to now, a regular ANN performance over completely different studies has not been achieved. Many factors contribute to the inconsistency in the performance of neural network models. One such factor is that ANN modeling involves determining a massive number of design parameters, and the current style practice is basically heuristic and circumstantial, this does not exploit the full potential of neural networks. Systematic ANN modeling processes and ways for TSF are, thus, greatly required. Motivated by this would like, this paper tries to develop an automatic ANN modeling theme. It's primarily based on the generalized regression neural network (GRNN), a special sort of neural network. By benefiting from several GRNN properties (i.e., one design parameter and fast learning) and by incorporating many design ways (e.g., fusing multiple GRNNs), we have a tendency to are able to create the proposed modeling theme to be effective for modeling massive-scale business time series. The initial model was entered into the NN3 time-series competition. It was awarded the most effective prediction on the reduced dataset among approximately 60 totally different models submitted by scholars worldwide.