A learning-based strategy for portfolio selection

Neural networks have shown exceptional performance in targeting different research areas. In this paper, we investigate a learning-based strategy for optimal investment by using neural network. First, an optimization problem for portfolio selection is proposed. Then, a neural network model is used to optimize this problem. The main contribution is that based on this proposed optimization problem and neural network model, we can easily implement the structure and obtain the final results by using deep learning software. Finally, we numerically compare the results obtained from our strategy with those of classic solutions. The comparison demonstrates the effectiveness of the learning-based strategy.

Automated summary

This research investigates a learning-based strategy for optimal investment using neural networks. By proposing an optimization problem and utilizing a neural network model, the structure can be easily implemented and final results can be obtained through deep learning software. Numerical comparison with classic solutions demonstrates the effectiveness of the learning-based strategy.