A Triangular Network For Density Estimation
This work provides a modular and efficient density estimation method for high-dimensional data, but it appears incremental as it builds on existing neural autoregressive flow techniques.
The authors tackled the problem of density estimation for high-dimensional data by implementing a triangular neural network based on neural autoregressive flow, achieving state-of-the-art bits-per-dimension indices of about 1.1 on MNIST and 3.7 on CIFAR-10.
We report a triangular neural network implementation of neural autoregressive flow (NAF). Unlike many universal autoregressive density models, our design is highly modular, parameter economy, computationally efficient, and applicable to density estimation of data with high dimensions. It achieves state-of-the-art bits-per-dimension indices on MNIST and CIFAR-10 (about 1.1 and 3.7, respectively) in the category of general-purpose density estimators.