Abstract | ||
---|---|---|
We present an efficient algorithm for maximum likelihood estimation (MLE) of exponential family models, with a general parametrization of the energy function that includes neural networks. We exploit the primal-dual view of the MLE with a kinetics augmented model to obtain an estimate associated with an adversarial dual sampler. To represent this sampler, we introduce a novel neural architecture, dynamics embedding, that generalizes Hamiltonian Monte-Carlo (HMC). The proposed approach inherits the flexibility of HMC while enabling tractable entropy estimation for the augmented model. By learning both a dual sampler and the primal model simultaneously, and sharing parameters between them, we obviate the requirement to design a separate sampling procedure once the model has been trained, leading to more effective learning. We show that many existing estimators, such as contrastive divergence, pseudo/composite-likelihood, score matching, minimum Stein discrepancy estimator, non-local contrastive objectives, noise-contrastive estimation, and minimum probability flow, are special cases of the proposed approach, each expressed by a different (fixed) dual sampler. An empirical investigation shows that adapting the sampler during MLE can significantly improve on state-of-the-art estimators(1). |
Year | Venue | Keywords |
---|---|---|
2019 | ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 32 (NIPS 2019) | maximum likelihood estimation (mle),neural networks,noise-contrastive estimation |
Field | DocType | Volume |
Entropy estimation,Mathematical optimization,Embedding,Parametrization,Hamiltonian (quantum mechanics),Exponential family,Algorithm,Artificial neural network,Mathematics,Adversarial system,Estimator | Journal | 32 |
ISSN | Citations | PageRank |
1049-5258 | 0 | 0.34 |
References | Authors | |
0 | 7 |
Name | Order | Citations | PageRank |
---|---|---|---|
Bo Dai | 1 | 230 | 34.71 |
Zhen Liu | 2 | 40 | 5.01 |
Hanjun Dai | 3 | 323 | 25.71 |
Niao He | 4 | 212 | 16.52 |
Arthur Gretton | 5 | 3638 | 226.18 |
Le Song | 6 | 2437 | 159.27 |
Dale Schuurmans | 7 | 2760 | 317.49 |