Cell-expanded Long Short-term Memory

Long short-term memory (LSTM) is a recurrent neural network (RNN) that is widely used for historical time series prediction. LSTM has a memory architecture called a memory cell, which enables long-term time-series prediction. Many historical time series have complex changes; thus, a single memory cell cannot fully capture the features. In this paper, we propose a method that exhibits good predictive performance against complex time-series changes by incorporating additional memory cells.