This paper presents a cooperative coevolutionary approach for stock trading model using Genetic Network Programming-Sarsa called CCGNP-Sarsa. Although theoretically, a single algorithm with sufficient size could solve any problem, in practice the stock market problem is too large and too complex to construct the appropriate algorithm to solve it. For such problems, cooperative coevolution which simultaneously evolves several species with the sum of their fitness values has been proposed as a successful alternative and was applied to make the stock trading models an integrated one. Such an approach allows different species of the GNP-Sarsa model to evolve in a parallel and cooperative manner, which makes the generated model more robust, generalized and efficient for generating stock trading strategies. CCGNP-Sarsa places as few restrictions as possible to the structure, allowing the model to obtain a wide variety of architecture during the evolution and to be easily used to solve complicated problems. To confirm the effectiveness of the proposed method, the simulations are carried out and compared with other methods like GNP-Sarsa with subroutines, GNP-Sarsa and Buy&Hold method. The results shows that the stock trading models using CCGNP-Sarsa outperforms all the other methods.
A Comparative Evaluation of Predominant Deep Learning Quantified Stock Trading Strategies
