Programming II

1.) KNOWLEDGE: - To know the basics of extracting knowledge from data - To know the different tasks that can be solved with machine learning - To know the different techniques of machine learning and their characteristics - To know the methodology of knowledge extraction and the phases involved - To know tools available for extracting knowledge 2.) UNDERSTANDING: - To understand the basic concepts of knowledge extraction - To understand the basics and motivations of data mining - To understand the methodology and the different phases of knowledge extraction - To understand the usefulness of different techniques for extracting knowledge - To understand the differences of different representations: propositional and relational - To understand the relationship between model complexity, amount of data, characteristics of the problem and overfitting 3.) APPLICATION: - Analyze the domain and design knowledge extraction processes adapted to the problem. - Evaluate the performance and efficiency of different methods of extracting knowledge - Work on specific domains and compare different techniques to verify their performance in extracting knowledge 4.) EVALUATION - Selection of algorithms, model selection and parameter adjustment. - Consider the relationship between computational cost and marginal improvement of different solutions - Assessment of whether the results are adequate, compared to random or basic algorithms

1. Introduction to Machine Learning 2. Basic methods for classification and regression: 2.1. Nearest neighbour (KNN) 2.2. Trees and Rules 3. Machine Learning pipeline 3.1. Training 3.2. Hyper-parameter tuning 3.3. Evaluation 3.4. Preprocessing and feature selection 4. Machine learning with R and with the MLR (mlr3) library. 5. Advanced methods for classification and regression 5.1. Ensembles: bagging, random forests, boosting 5.2. Support Vector Machines 6. Brief introduction to machine learning in Python (scikit-learn library)

Theory: Lectures will be focused on teaching all concepts related to machine learning. They will be carried out in synchronous on-line mode. Practical computer Sessions: The practical classes will be developed so that, in a supervised way, students learn to solve real problems with machine learning. The practices will be carried out in groups of 2 students. There are several assignments related to topics in the course. There will be tutorials to help the understanding both of theory and practice.