CHOOSING A HARDWARE PLATFORM FOR MACHINE LEARNING

Perepelkin Vadim

Despite the fact that the basic principles of neural networks were formulated by Warren McCulloch and Walter Pitts back in 1947, today's time can be called as the era of the development of machine learning and artificial intelligence without exaggeration. It is over the past few years that there has been a breakthrough in the development of neural network technologies and their industrial applications. Neural networks have been successfully used in computer vision, natural language processing, building autopilot systems, robotics, etc. To a large extent, all this became possible due to the development of computer technology and the appearance of new hardware platforms that show high performance when handle mathematical operations, which also facilitated to create and train deep neural networks with a large number of layers and neurons. The problem areas of training neural networks and the efficiency of using various hardware platforms by an example of training LeNet-5 neural network with MNIST dataset are considered at the research work.

Keywords:machine learning, neural networks, hardware platforms, processors.