LSTM Rollout Curriculum Using Double Pendulum
International Conference on Artificial Intelligence 2024
Reinis Freibergs, Ēvalds Urtāns, Ansis Ecis, Henrik Gabrielyan

In this work we model a double pendulum system with deep neural networks based on a dataset generated from video recordings. For comparison, a similar model is made by describing the system with differential equations. Actually compared are both models’ capabilities of predicting the next 2s of double pendulum motion by using information about the previous second. In addition, both models are compared by their ability to make predictions in specific error margin. Results show that deep learning based approaches give much better predictions, where the best deep learning based model could predict the next 1.5s in a specified error margin, while the best differential equation based one only 0.12s, all other metrics agree with this result as well.


DOI
10.18178/ijml.2024.14.1.1151
Hipersaite
https://www.ijml.org/show-133-1356-1.html

Freibergs, R., Urtāns, Ē., Ecis, A., Gabrielyan, H. LSTM Rollout Curriculum Using Double Pendulum. International Conference on Artificial Intelligence, 2024, Vol. 14, No. 1, 12.-17.lpp. ISSN 2972-368X. Pieejams: doi:10.18178/ijml.2024.14.1.1151

Publikācijas valoda
English (en)
RTU Zinātniskā bibliotēka.
E-pasts: uzzinas@rtu.lv; Tālr: +371 28399196