The course introduces basic approaches to Machine Learning and Deep Learning and classical methods used in the field. Practical use of the methods is demonstrated on solving simple engineering problems.
- Project 1 - ANN Clasifier
- Project 2 - Adaptive ANN model for function approximation
Simple ANN Clasifier. Inputs are coordinates of point x. Output is binary value True/False whether the point is inside or outside of defined ellipse.
Instead of using libraries for machine learning, simple ANN was created using Numpy library.
Adaptive ANN modelling for Rastrigin and Schwefel functions. Nelder-Mead Optimization was used. For this project, Keras API from Tensorflow framework was used.
