Stock Trading Strategies Based on Deep Reinforcement Learning

Figure 1

The overall structure of the model. represents candlestick chart feature, represents the feature of stock data and technical indicators, and the feature vector obtained by contacting these two feature vectors is used as the input of the two fully connected (FC) layers. In this paper, FC layers are used to construct the dueling DQN network; the two FC layers represent the advantage function and state value function in dueling DQN. The final Q value is obtained by adding the outputs of the two functions.