Title: Deep learning approach for forecasting financial time series trends Authors:  Yosi Keller - bar ilan university (Israel) [presenting]
Abstract: A data-driven end-to-end Deep Learning approach is presented for time series prediction, applied to financial time series. A Deep Learning scheme is derived to predict the temporal trends of stocks and ETFs in NYSE or NASDAQ. Our approach utilizes a neural network (NN) applied to raw financial data inputs and is trained to predict the temporal trends of stocks and ETFs. To handle commission-based trading, an investment strategy is derived that utilizes the probabilistic outputs of the NN and optimizes the average return. The proposed scheme is experimentally shown to provide statistically significant accurate predictions of financial market trends. The investment strategy is shown to be profitable in realistic commissions-based trading while comparing favourably with contemporary benchmarks over two years of back-testing.