Multi-Class Assessment Based on Random Forests

Today, many students are moving towards higher education courses that do not suit them and end up failing. The purpose of this study is to help provide counselors with better knowledge so they can offer future corresponding their profile. second objective allow the teaching staff propose training adapted by anticipating possible difficulties. This thanks a machine learning algorithm called Random Forest, allowing for classification depending on results. We had process data, generate models using our algorithm, cross results obtained have final prediction. tested method different use cases, from two classes five classes. These sets represent intervals an average ranging 0 20. Thus, accuracy 75% was achieved set 85% three