Recognition of the MNIST dataset with defective rows
Аrtеm N. Shevlyakov, A A Berezin · Journal of Physics Conference Series · 2022
Abstract One of the well-known classification problems in machine learning is the problem of recognizing handwritten numbers. This task is solved, since there are different neural networks that determine with fairly high accuracy which number is shown in the picture. In this paper, we will consider the recognition of handwritten digits subjected to certain deformations.