#Machine Learning Project

Machine Learning (ML) is a field of artificial intelligence (AI) that provides the systems with the ability to automatically learn and improve from experience without being explicitly programmed. ML computer programs focus on a way to access data and use them to learn for themselves. The process of learning begins with observing data, in order to look for patterns and make better decisions in the future based on the examples provided. The main goal is to allow computers to learn automatically without human intervention or assistance. Learning algorithms are generally based on the optimization of a performance criterion that measures how well the selected “machine” matches the available data . The objective of this project is to study different machine learning algorithms dedicated to the classification of two different datasets: Banknote Authentication and Chronic Kidney Disease. However, the main goal is not the programming part itself, but the analysis of their behaviour in each dataset and the extraction of different conclusions on their performances. Since both datasets are examples of supervised learning, where the instances are given with known labels (i.e., the corresponding correct outputs), the clustering algorithm, which is used for unsupervised learning, was discarded. The goal of supervised learning is to build a concise model of the distribution of class labels in terms of predictor features. The resulting classifier is then used to assign class labels to the testing instances where the values of the predictor features are known, but the value of the class label is unknown . In particular, this work is concerned with classification problems, in which the output of instances admits only discrete values. It was decided to study Principal Component Analysis (PCA), Support Vector Machine (SVM), Neural Networks and Decision Trees. These four different algorithms represent the most popular ways of classifying data in supervised machine learning. As stated before, in this project two datasets need to be analysed. The Banknote Authentication dataset consists of 1372 instances: 762 are non authentic notes and 610 are authentic. The features are five and represents the variance, the skewness, the curtosis, the entropy and the class (either authentic notes or not) .