Deep reinforcement learning for time series: playing idealized trading games

Explore this paper's citation graph

Summary

Deep Q-learning is investigated as an end-to-end solution to estimate the optimal strategies for acting on time series input to test whether the agent can capture the underlying dynamics and utilize the hidden relation among the inputs.

Type
preprint
Published
2018-03-11
Cited by
16
References
21
Access
Open access

Keywords

Univariate, Bivariate analysis, Reinforcement learning, Series (stratigraphy), Computer science

References

Cited by

Related papers