Abstract
The brain structures are key indicators to represent the complexity of many cognitive functions, e.g., visual pathways and memory circuits. Inspired by the topology of the mouse brain provided by the Allen Brain Institute, whereby 213 brain regions are linked as a mesoscale connectome, we propose a mouse-brain topology improved evolutionary neural network (MT-ENN). The MT-ENN model incorporates parts of biologically plausible brain structures after hierarchical clustering, and then is tuned by the evolutionary learning algorithm. Two benchmark Open-AI Mujoco tasks were used to test the performance of the proposed algorithm, and the experimental results showed that the proposed MT-ENN was not only sparser (containing only 61% of all connections), but also performed better than other algorithms, including the ENN using a random network, standard long-short-term memory (LSTM), and multi-layer perception (MLP). We think the biologically plausible structures might contribute more to the further development of artificial neural networks.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Doya, K.: Reinforcement learning in continuous time and space. Neural Comput. 12(1), 219–245 (2000)
Hasani, R., Lechner, M., Amini, A., Rus, D., Grosu, R.: Liquid time-constant networks. arXiv preprint arXiv:2006.04439 (2020)
Hasani, R., Lechner, M., Amini, A., Rus, D., Grosu, R.: A natural lottery ticket winner: reinforcement learning with ordinary neural circuits. In: International Conference on Machine Learning, pp. 4082–4093. PMLR (2020)
Hochreiter, S., Schmidhuber, J.: Long short-term memory. Neural Comput. 9(8), 1735–1780 (1997)
Oh, S.W., et al.: A mesoscale connectome of the mouse brain. Nature 508(7495), 207–214 (2014)
Rubinov, M., Ypma, R.J., Watson, C., Bullmore, E.T.: Wiring cost and topological participation of the mouse brain connectome. Proc. Natl. Acad. Sci. 112(32), 10032–10037 (2015)
Schulman, J., Wolski, F., Dhariwal, P., Radford, A., Klimov, O.: Proximal policy optimization algorithms. arXiv preprint arXiv:1707.06347 (2017)
Wildenberg, G.A., Rosen, M.R., Lundell, J., Paukner, D., Freedman, D.J., Kasthuri, N.: Primate neuronal connections are sparse in cortex as compared to mouse. Cell Rep. 36(11), 109709 (2021)
Zhang, D., Zhang, T., Jia, S., Xu, B.: Multiscale dynamic coding improved spiking actor network for reinforcement learning. In: Thirty-Sixth AAAI Conference on Artificial Intelligence (2022)
Zhang, T., Zeng, Y., Xu, B.: A computational approach towards the microscale mouse brain connectome from the mesoscale. J. Integr. Neurosci. 16(3), 291–306 (2017)
Zhang, T., et al.: Neuron type classification in rat brain based on integrative convolutional and tree-based recurrent neural networks. Sci. Rep. 11(1), 7291 (2021)
Acknowledgements
This work was supported in part by the Shanghai Municipal Science and Technology Major Project (2021SHZDZX), the Youth Innovation Promotion Association CAS, and the National Natural Science Foundation of China under Grants No. 82171992 and 81871394.
Author information
Authors and Affiliations
Corresponding authors
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2022 IFIP International Federation for Information Processing
About this paper
Cite this paper
Han, X., Jia, K., Zhang, T. (2022). Mouse-Brain Topology Improved Evolutionary Neural Network for Efficient Reinforcement Learning. In: Shi, Z., Jin, Y., Zhang, X. (eds) Intelligence Science IV. ICIS 2022. IFIP Advances in Information and Communication Technology, vol 659. Springer, Cham. https://doi.org/10.1007/978-3-031-14903-0_1
Download citation
DOI: https://doi.org/10.1007/978-3-031-14903-0_1
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-14902-3
Online ISBN: 978-3-031-14903-0
eBook Packages: Computer ScienceComputer Science (R0)
