Browse > Article
http://dx.doi.org/10.7236/JIIBC.2021.21.6.117

Comparison of Activation Functions using Deep Reinforcement Learning for Autonomous Driving on Intersection  

Lee, Dongcheul (Dept. of Multimedia Engineering, Hannam University)
Publication Information
The Journal of the Institute of Internet, Broadcasting and Communication / v.21, no.6, 2021 , pp. 117-122 More about this Journal
Abstract
Autonomous driving allows cars to drive without people and is being studied very actively thanks to the recent development of artificial intelligence technology. Among artificial intelligence technologies, deep reinforcement learning is used most effectively. Deep reinforcement learning requires us to build a neural network using an appropriate activation function. So far, many activation functions have been suggested, but different performances have been shown depending on the field of application. This paper compares and evaluates the performance of which activation function is effective when using deep reinforcement learning to learn autonomous driving on highways. To this end, the performance metrics to be used in the evaluation were defined and the values of the metrics according to each activation function were compared in graphs. As a result, when Mish was used, the reward was higher on average than other activation functions, and the difference from the activation function with the lowest reward was 9.8%.
Keywords
Autonomous Driving; Deep Learning; Reinforcement Learning;
Citations & Related Records
연도 인용수 순위
  • Reference
1 M. Kim, S. Lee, J. Lim, J. Choi, S. Kang, "Unexpected collision avoidance driving strategy using deep reinforcement learning", IEEE Access, Vol. 8, pp. 17243-17252, 2020. DOI: https://doi.org/10.1109/ACCESS.2020.2967509   DOI
2 L. Edouard, "An Environment for Autonomous Driving Decision-Making", GitHub, https://github.com/eleurent/highway-env, 2018.
3 A. Rasouli, J. K. Tsotsos, "Autonomous vehicles that interact with pedestrians: A survey of theory and practice", IEEE Trans. Intell. Transp. Syst., Vol. 21, No. 3, pp. 900-918, 2020. DOI: https://doi.org/10.1109/TITS.2019.2901817   DOI
4 P. Hart, A. Knoll, "Using counterfactual reasoning and reinforcement learning for decision-making in autonomous driving", arXiv:2003.11919, 2020.
5 G. Shu, W. Liu, X. Zheng, J. Li, "IF-CNN: Image-Aware Inference Framework for CNN With the Collaboration of Mobile Devices and Cloud", IEEE Access, Vol 6, pp. 68621-68633, 2018. DOI: https://doi.org/10.1109/ACCESS.2018.2880196   DOI
6 M. Lau, K. Lim, "Review of Adaptive Activation Function in Deep Neural Network", 2018 IEEE-EMBS Conference on Biomedical Engineering and Sciences, pp. 686-690, 2018. DOI: https://doi.org/10.1109/IECBES.2018.8626714.   DOI
7 S. Elfwing, E. Uchibe, K. Doya. "Sigmoid-weighted linear units for neural network function approximation in reinforcement learning", Neural Networks, Vol. 107, pp. 3-11, 2018. DOI: https://doi.org/10.1016/j.neunet.2017.12.012   DOI
8 E. Jang, J. Kim, "Proposal of New Information Processing Model for Implementation of Autonomous Mobile System", The Journal of The Institute of Internet, Broadcasting and Communication, Vol. 19, No. 2, pp. 237-242, 2019. DOI: https://doi.org/10.7236/JIIBC.2019.19.2.237   DOI
9 T. Jiang, J. Cheng, "Target Recognition Based on CNN with LeakyReLU and PReLU Activation Functions", 2019 International Conference on Sensing, Diagnostics, Prognostics, and Control, pp. 718-722, 2019. DOI: https://doi.org/10.1109/SDPC.2019.00136   DOI
10 D. Hendrycks, K. Gimpel, "Gaussian Error Linear Units", arXiv:1606.08415, 2020.
11 D. Misra, "Mish: A Self Regularized Non-Monotonic Activation Function", arXiv:1908.08681, 2020.
12 J. Schulman, F. Wolski, P. Dhariwal, A. Radford, O. Klimov, "Proximal Policy Optimization Algorithms", arXiv:1707.06347, 2017.