The purpose of these lectures today is to review a few rather basic Machine Learning algorithms, while trying to see them from a Data Mining perspective. Thus, we will discuss the very notion of modelling, its role within the process of Knowledge Discovery from Data, and some of the particularities of this specific context. We will go through two "descriptive modelling" processes, namely k-means clustering and association rule mining; we will discuss some generalities about "predictive modelling", such as ROC-based evaluation and the bias-variance trade-off, and discuss some specific simple classifiers: naïve Bayes, nearest neighbours, linear classifiers, their extension using kernels, and the Adaboost metapredictor.