Forecasting Directional Movement of Stock Prices using Deep Learning

Stock market’s volatile and complex nature makes it difficult to predict the market situation. Deep Learning is capable of simulating analyzing patterns in unstructured data. learning models have applications image recognition, speech natural language processing (NLP), many more. Its application stock prediction gaining attention because its capacity handle large datasets data mapping with accurate prediction. However, most methods ignore impact mass media on company’s investors’ behaviours. This work proposes a hybrid deep model combining Word2Vec long short-term memory (LSTM) algorithms. The main objective design an intelligent tool forecast directional movement prices based financial time series news headlines as inputs. binary predicted output obtained using proposed would aid investors making better decisions. effectiveness assessed terms accuracy five companies from different sectors operation.