Digital Data Sciences

Notions et techniques abordées

Machine Learning Basics - Generalization, Capacity, Overfitting and Underfitting - Trading off Bias and Variance and the Mean Squared Error - Consistency - Supervised learning - Unsupervised learning - Data Understanding - Attribute Understanding - Data Quality - Correlation Analysis - Outlier Detection - Missing Values - Principles of Modeling - Model Classes - Fitting Criteria and Score Functions - Error Functions for Classification Problems - Gradient Method - Types of Errors - Model Validation - Cross-Validation - Bootstrapping - Measures for Model Complexity - Data Preparation - Selecting Data - Feature Selection - Dimensionality Reduction - Clean Data - Construct Data - Complex Data Types - Finding Patterns - Notion of (Dis- )Similarity - Hierarchical Clustering - Prototype and Model-Based Clustering - Density-Based Clustering - Finding Explanations - Decision Trees - Finding Predictors - Nearest-Neighbor Predictors - Artificial Neural Networks - Support Vector Machines - Ensemble Methods -Combining Models - Committees - Boosting - Mixtures of experts - Evaluation and Deployment.