FenixEdu™

AT1 - Data Science

Aula 1

Data science: context and goals.
Course organization and planning.

AT2 - KDD Process

Aula 2

Basic concepts: data, records, attributes and information.
KDD process: from data to information.

AT3 - Deloitte presentation

Aula 3

Case study presentation by Deloitte.

AT3 - Data profiling

Aula 4

Data profiling: granularity, distribution, dimensionality and sparsity.

AT5 - Modelling

Aula 5

Modelling: mining tasks; supervised, semi-supervised and unsupervised learning.

Evaluation principles and the Occam's razor.

AT6 - Classification

Aula 6

Classification. Notion of Concept.

Evaluation: measures, training strategies and statistical significance.

AT7 - Classification: analogizers

Aula 7

Analogizers: KNN algorithm and the idea behind Support Vector Machines.

Data preparation: scaling, distance measures and dummification

AT8 - Classification: Bayesians

Aula 8

Bayesians: MAP and Naive Bayes. Other Bayesian approaches.

Data. preparation: missing value imputation.

AT9 - Classification: symbolists

Aula 9

Symbolists: decision trees - algorithms, measures.
Overfitting and the pruning.

AT10 - Classification: ensembles - bagging

Aula 10

Ensembles: bagging and random forests.
Feature engineering: selection and generation.

AT11 - Classification: evolutionists

Aula 11

Evolutionists: genetic algorithms - introduction and motivation.
Data preparation: balancing - resampling and SMOTE.
Evaluation: other measures and ROC charts.

AT12 - Classification: connectionists

Aula 12

Neural networks: main idea and gradient descent algorithm.

Deep learning - a brief summary.

AT13 - Regression

Aula 13

Regression: linear and logistic.

AT14 - Classification: ensemble - boosting

Aula 14

Boosting and gradient boosting algorithm.

AT15 - Clustering

Aula 15

Introduction to clustering data analysis.

Clustering approaches: partitioning, hierarchical, density-based and model-based.

Evaluation criteria: cohesion and compactness.

AT16 - Feature extraction

Aula 16

Feature extraction: Principal component analysis (PCA) and linear discriminant analysis (LDA).

AT17 - Pattern mining

Aula 17

Pattern Mining. Apriori algorithm and other approaches.

Evaluation: support, confidence and lift.

AT18 - Sequential pattern mining

Aula 18

Sequential Pattern Mining.
Discretization methods.

AT19 - Anomaly detection

Aula 19

Anomaly detection: main approaches and applications.

AT20 - Time series

Aula 20

Time series: representations - SAX, discrete Fourier and wavelet transforms.

Measuring similarity among time series - dynamic time warping.

AT21 - Mining time series

Aula 21

Forecasting: time series decomposition, description and methods.

Other mining tasks: the matrix profile based approaches.

AT22 - Network data analysis

Aula 22

Network data analysis: description and properties.

AT23 - Social and web data analysis

Aula 23

SNA: HITS and PageRank algorithms.

AT24 - Big data

Aula 24

Big data: challenges.

Data streams and large scale data mining.

Parallelisation.

AT26 - Privacy and ethical concerns

Aula 25

Privacy issues, ethical concerns and the GDPR.
Anonymization.

AT27 - Closing remarks

Aula 26

Data science and main challenges.

Closing remarks.

Enrollment

Aula 1

Group enrolment - no lab session.

Lab 0 (Week 2)

Aula 2

Project schedule and resource allocation - non graded

Lab 1 (Week 3)

Aula 3

Data profiling - dimensionality, distribution and granularity

Lab 2 (Week 4)

Aula 4

Data profiling - sparsity and correlation

Lab 3 (Week 5)

Aula 5

Naive Bayes and KNN

Lab 4 (Week 6)