Hello everyone with an interest in The Information Age. I will announce today that I’ll start to post regular reviews of papers that I deem interesting developments in fields of research this blog cares about, namely to do with Information Technology, Computer Science and Data Science.
My first choice for today is a paper on the thriving and important topic of Neural Networks, Deep Neural Networks and Evolutionary Computing, topics that are undergoing a period of massive improvements and feeding the Artificial Intelligence resurgence.
I describe here today a project called Horn and its newest development around large-scale neural networks and the performance in regularization and training of such networks. Experiments were performed on MNIST handwritten digits classification including results, as depicted in the abstract:
I introduce a new distributed system for effective training and regularizing of Large-Scale Neural Networks on distributed computing architectures. The experiments demonstrate the effectiveness of flexible model partitioning and parallelization strategies based on neuron-centric computation model, with an implementation of the collective and parallel dropout neural networks training. Experiments are performed on MNIST handwritten digits classification including results.