Mixtures of Sparse Autoregressive Networks
This work addresses distribution estimation for high-dimensional data, offering a scalable and efficient method, though it appears incremental as it builds on existing concepts like sparsity and mixtures.
The authors tackled high-dimensional distribution estimation by developing a mixture of sparse autoregressive networks, achieving state-of-the-art or better results on standard benchmarks with fast training and excellent generalization.
We consider high-dimensional distribution estimation through autoregressive networks. By combining the concepts of sparsity, mixtures and parameter sharing we obtain a simple model which is fast to train and which achieves state-of-the-art or better results on several standard benchmark datasets. Specifically, we use an L1-penalty to regularize the conditional distributions and introduce a procedure for automatic parameter sharing between mixture components. Moreover, we propose a simple distributed representation which permits exact likelihood evaluations since the latent variables are interleaved with the observable variables and can be easily integrated out. Our model achieves excellent generalization performance and scales well to extremely high dimensions.