Machine Learning

Recommended Prerequisites

Probabilities and Statistics, Linear Algebra, Calculus, Programming, Machine Learning - I

Teaching Methods

The course consists of theoretical classes (T) that provide a detailed presentation of concepts, principles, and fundamental theories using audiovisual aids. Basic practical exercises are solved during these classes to demonstrate the practical relevance of the subject and illustrate its application to real-life scenarios. Additionally, practical laboratory classes (PL) are conducted, where students work on exercises that require the integration of various theoretical concepts and foster critical reasoning skills.

The evaluation process encompasses all the topics covered in the course and focuses on assessing understanding of the fundamental theoretical concepts as well as the ability to solve complex problems.

Learning Outcomes

The course focuses on teaching algorithms that enable the exploration of machine learning applications, both supervised and unsupervised, for classification and regression tasks. It aims to formalize these algorithms and provide models that are applicable across various domains. This course is a continuation of Machine Learning-I, with the goal of delving deeper into previously covered topics, exploring new concepts, and introducing advanced techniques related to classification and regression.

Work Placement(s)

No

Syllabus

1. Introduction

i.          Pipeline review

ii.         Concepts about supervised and unsupervised models

2-Unsupervised learning

i.          Partition algorithms: k-Means and k-Mendoids

ii.         Hierarchical Clustering: Agglomerative and Divisive

iii.         Density-based clustering: DBSCAN

iv.        Clustering based on combined Gaussian models

v.         Assessment methods: Intrinsic and extrinsic

3-Supervised non-connectionist models

i.          Linear (Euclidian and Mahalanobis) and Fisher Discriminants

ii.         Bayes classification; Bayes Estimation and Risk; Maximum A Posteriori (MAP).

iii.         K Nearest Neighbor Algorithm (kNN)

iv.        Support Vector Machines (SVM)

v.         Combination of Classifiers: Product and Sum Rule

4-Connectionist Supervised Models

i.          Non-recurrent neural networks: MLP and RBF

ii.         Recurrent neural networks: Vanilla, GRU, LSTM

5-Rule-based models

i.          Decision trees

6- Complementary aspects

i.  semi-supervised learning, reinforcement learning.

Assessment Methods

Assessment
Project: 40.0%
Exam: 60.0%

Bibliography

1. Bishop, C.M., ―Pattern Recognition and Machine Learning‖, Springer Verlag, 2006.

2. Duda, R. O., Hart, P.E., and Stork, D.G., ―Pattern Classification, ‖ 2nd ed. Wiley Interscience (2001).

3. J.P. Marques de Sá, ―Pattern Recognition: Concepts, Methods and Applications‖, 2001, XIX, 318 p., 197 illus., Springer-Verlag (2001).

4. M. N. Murty and V. S. Devi, ―Pattern Recognition: An Algorithmic Approach, Springer, 1st Edition., XII, 263 p. (2011).

5. Peter Flach, Machine Learning: the art and science of algorithms that make sense of data, Cambridge University Press, 2012.

6. Trevor Hastie, Robert Tibsshirani and Jerome Friedman, The Elements of Statistical Learning (2nd Edition), Springer, 2008.

7. Introduction to Machine Learning with Python, Andreas C. Muller and Sarah Guido, O'Reilly, 2017.

8. João Gama et alli, Extracção de Conhecimento de Dados, Edições Sílabo, 2012.