Abstract | ||
---|---|---|
High-dimensional observations and unknown dynamics are major challenges when applying optimal control to many real-world decision making tasks. The Learning Controllable Embedding (LCE) framework addresses these challenges by embedding the observations into a lower dimensional latent space, estimating the latent dynamics, and then performing control directly in the latent space. To ensure the learned latent dynamics are predictive of next-observations, all existing LCE approaches decode back into the observation space and explicitly perform next-observation prediction---a challenging high-dimensional task that furthermore introduces a large number of nuisance parameters (i.e., the decoder) which are discarded during control. In this paper, we propose a novel information-theoretic LCE approach and show theoretically that explicit next-observation prediction can be replaced with predictive coding. We then use predictive coding to develop a decoder-free LCE model whose latent dynamics are amenable to locally-linear control. Extensive experiments on benchmark tasks show that our model reliably learns a controllable latent space that leads to superior performance when compared with state-of-the-art LCE baselines. |
Year | Venue | DocType |
---|---|---|
2020 | ICML | Conference |
Citations | PageRank | References |
0 | 0.34 | 0 |
Authors | ||
8 |
Name | Order | Citations | PageRank |
---|---|---|---|
Rui Shu | 1 | 32 | 9.88 |
Nguyen Tung | 2 | 0 | 0.34 |
Chow, Yinlam | 3 | 98 | 14.03 |
Tuan Pham | 4 | 503 | 73.75 |
Than Khoat | 5 | 0 | 0.34 |
Mohammad Ghavamzadeh | 6 | 814 | 67.73 |
Stefano Ermon | 7 | 726 | 78.25 |
Hung Hai Bui | 8 | 1188 | 112.37 |