Optimizing Training Trajectories in Variational Autoencoders via Latent Bayesian Optimization Approach
This work addresses the inefficiency of hyperparameter tuning for VAEs in physics and materials science, though it appears incremental as it adapts existing optimization methods to a specific model.
The paper tackled the problem of hyperparameter trajectory optimization in variational autoencoders (VAEs) by proposing a latent Bayesian optimization (zBO) approach, which demonstrated improved performance in finding joint discrete and continuous rotationally invariant representations for datasets like MNIST and experimental plasmonic nanoparticle data.
Unsupervised and semi-supervised ML methods such as variational autoencoders (VAE) have become widely adopted across multiple areas of physics, chemistry, and materials sciences due to their capability in disentangling representations and ability to find latent manifolds for classification and regression of complex experimental data. Like other ML problems, VAEs require hyperparameter tuning, e.g., balancing the Kullback Leibler (KL) and reconstruction terms. However, the training process and resulting manifold topology and connectivity depend not only on hyperparameters, but also their evolution during training. Because of the inefficiency of exhaustive search in a high-dimensional hyperparameter space for the expensive to train models, here we explored a latent Bayesian optimization (zBO) approach for the hyperparameter trajectory optimization for the unsupervised and semi-supervised ML and demonstrate for joint-VAE with rotational invariances. We demonstrate an application of this method for finding joint discrete and continuous rotationally invariant representations for MNIST and experimental data of a plasmonic nanoparticles material system. The performance of the proposed approach has been discussed extensively, where it allows for any high dimensional hyperparameter tuning or trajectory optimization of other ML models.