Training Stacked Denoising Autoencoders for Representation Learning
This work addresses representation learning for image classification, but appears incremental as it builds on existing autoencoder methods with a new optimization approach.
The paper tackled the problem of learning powerful representations from high-dimensional data using stacked denoising autoencoders, by implementing and comparing stochastic gradient descent and a novel genetic algorithm-based approach, resulting in performance analysis on standard image classification datasets without specifying concrete numbers.
We implement stacked denoising autoencoders, a class of neural networks that are capable of learning powerful representations of high dimensional data. We describe stochastic gradient descent for unsupervised training of autoencoders, as well as a novel genetic algorithm based approach that makes use of gradient information. We analyze the performance of both optimization algorithms and also the representation learning ability of the autoencoder when it is trained on standard image classification datasets.