Trading strategy model based on LSTM neural network and Extreme Value-Dynamic programming

The advantage of a trading strategy is that it can help us identify potential trading opportunities. Traders need to use the best investment strategy to take into account factors such as market volatility and transaction costs, so as to get the maximum benefit on the target date. this paper uses LSTM neural network to make predictions. The predicted value is the price of gold and bitcoin five days after the investment date, the model fits well. By analyzing the problem, the goal is to find the optimal value, which can be regarded as a multi-stage decision-making process optimization problem. Therefore, this paper selects the Extreme value-DP model for planning and decision-making, and uses the maximum (minimum) value to evaluate the strategy.