Demonstration of Neural Network in Prediction of Bearing Lifetime

The topic of this paper is the application of machine learning and neural networks in engineering, specifically in the prediction of the lifetime of bearings operating in different conditions. In addition, the basics of ma-chine learning are introduced, giving an idea of the importance of input data quality for model training. It also presents the elements of neural network training to be used in other projects. The article is supplemented by a source code examples written using only the Python language, and some other popular libraries, like the NumPy, Matplotlib, Tensorflow, Keras, and Scikit-learn. The main advantage of the libraries used is that they are freely available and widely used, bringing variety of sophisticated tools for gen-eral use.