Applications of deep learning in stock market prediction: Recent progress

Stock market prediction has been a classical yet challenging problem, with the attention from both economists and computer scientists. With purpose of building an effective model, linear machine learning tools have explored for past couple decades. Lately, deep models introduced as new frontiers this topic rapid development is too fast to catch up. Hence, our motivation survey give latest review recent works on stock prediction. We not only category different data sources, various neural network structures, common used evaluation metrics, but also implementation reproducibility. Our goal help interested researchers synchronize progress them easily reproduce previous studies baselines. Base summary, we highlight some future research directions in topic.