Measuring and Comparing the Performance of Evolutionary Algorithms for Automatic Machine Learning for Trend Prediction

Machine learning techniques and algorithms are used in almost every application domain such as image recognition, object detection, and financial applications. However, building a high-quality machine learning model requires vast amounts of human expertise. To solve this problem, researchers have looked towards automating machine learning techniques. Automatic Machine Learning problems often boil down to becoming an optimization problem, i.e. automatically optimizing the machine learning model without human professionals. In our project we explored a special class of nature inspired methods of optimization in order to solve this problem.