Deep Dive into Semi-Supervised ELBO for Improving Classification Performance
arXiv:2108.12734
Abstract
Decomposition of the evidence lower bound (ELBO) objective of VAE used for density estimation revealed the deficiency of VAE for representation learning and suggested ways to improve the model. In this paper, we investigate whether we can get similar insights by decomposing the ELBO for semi-supervised classification using VAE model. Specifically, we show that mutual information between input and class labels decreases during maximization of ELBO objective. We propose a method to address this issue. We also enforce cluster assumption to aid in classification. Experiments on a diverse datasets verify that our method can be used to improve the classification performance of existing VAE based semi-supervised models. Experiments also show that, this can be achieved without sacrificing the generative power of the model.
arXiv commentUnder Review
References in corpus (8)
- Adam: A Method for Stochastic Optimization
- Unsupervised and Semi-supervised Learning with Categorical Generative Adversarial Networks
- Auxiliary Deep Generative Models
- Ladder Variational Autoencoders
- Adversarial Variational Bayes: Unifying Variational Autoencoders and Generative Adversarial Networks
- Structured Disentangled Representations
- Graph Partition Neural Networks for Semi-Supervised Classification
- Distribution Matching in Variational Inference