Data Mining: Lecture Notes

Typeset lecture notes for Data Mining, University of Pisa, academic year 2024/25. The notes cover the two-part course: Data Mining I lays the foundations (data understanding, clustering, classification, regression, pattern and rule mining), and Data Mining II moves to advanced tabular and time-series methods (imbalanced learning, dimensionality reduction, outlier detection, gradient descent, logistic regression, SVM, neural and deep networks, ensembles, explainability, and the full time-series pipeline through similarity, clustering, and classification).

⚠️ Disclaimer. Derived from the Data Mining course materials (Academic Year 2024/2025), MSc in Data Science & Business Informatics, University of Pisa.

These notes are open educational content created by a student. They are not an academic source and may contain inaccuracies. You may freely share, modify, and reuse this material for educational and non-commercial purposes with appropriate attribution. The content is my personal interpretation of the professor's course materials and should not replace official teaching resources. I assume no responsibility for any errors or misinterpretations.

These notes were produced with an AI-in-the-middle workflow: a first human pass, then Claude Code to support formulation, understanding, and rewriting, followed by a final human review.

If you find errors, have suggestions, or spot unintentionally included copyrighted material (which I will promptly remove on notification), contact me at sclfnc@proton.me.

This course is part of the MSc Data Science lecture notes collection (University of Pisa), one repository per course. Clone the whole set with git clone --recursive.

Layout

• main.tex: entry point, in the folder root; identical across the whole notes collection (it only loads the shared preamble and the course file).
• src/housestyle.tex, shared house style: geometry, colors, section and ToC formatting, running heads, and the math environments (theorem, definition, …).
• src/common-preamble.tex: the shared package set, identical across courses.
• src/course.tex, everything specific to this course: title metadata, the two \part groups, custom itemize/enumerate bullets, the Rcode listings environment (syntax-highlighted R snippets), hyperref, and the \input{sec/...} list.
• sec/*.tex: section files (the body of the notes), pulled in by course.tex; named A_*–F_* for Data Mining I and 00_*–14_* for Data Mining II.
• img/: bitmap figures (lecture-slide screenshots and hand-drawn photos) included via \includegraphics; most other diagrams are native TikZ.
• dm-notes.pdf: the compiled notes, in the folder root.
• No bibliography: the notes carry no .bib file and use no \cite.

Build

Build from the folder root:

latexmk main.tex

latexmk runs pdflatex as many times as needed. The compiled PDF is named dm-notes.pdf; it and all auxiliary files (.aux, .log, .toc, .bcf, .bbl, …) land in the folder root and are git-ignored (listed in .gitignore). A .latexmkrc in the folder sets this up (it names the output via $jobname). To do it by hand instead:

pdflatex -jobname=dm-notes main && pdflatex -jobname=dm-notes main

The second pass resolves the table of contents and cross-references. Requires a standard TeX Live installation. Alternatively, upload the folder to Overleaf (New Project → Upload Project), set main.tex as the main document, and compile.

Credits

Written by Francesco Secoli, revised with the help of Claude Code: the course slides and lectures were transcribed and refined into LaTeX, then reworked into standalone notes. Based on the Data Mining course (a.y. 2024/25), University of Pisa. Contributions welcome: open an issue or a pull request.

Contents

The notes are twenty-one sections across two parts, in reading order:

#	Part	Section	Topics
1	DM I	Data manipulation and understanding	Data visualization, data preparation, normalization, similarity and distance measures
2	DM I	Clustering	Cluster validity, K-Means and variants (bisecting, X-Means), model-based EM, hierarchical linkage, density-based DBSCAN / OPTICS / HDBSCAN
3	DM I	Classification	K-NN, Naïve Bayes, model evaluation and comparison, decision trees (splitting measures, pruning, regression trees), model selection
4	DM I	Linear regression	Least squares, alternative fitting methods, model evaluation, extensions
5	DM I	Pattern and association rule mining	Association rules, Apriori, interestingness measures, FP-Growth, special frequent patterns
6	DM I	Rule-based models	Rule-set properties, direct and indirect rule extraction, model characteristics
7	DM II	Imbalanced learning	Undersampling (Tomek, ENN, CNN), oversampling (SMOTE, ADASYN), class weights and threshold adjustment
8	DM II	Dimensionality reduction	Feature selection (variance, univariate), feature projection (RSP, PCA, SVD, MDS, Sammon, ISOMAP, t-SNE)
9	DM II	Outlier detection	Naive (IQR, HBOS), statistical, deviation-, depth-, proximity-, clustering-, high-dimensional, ensemble- and model-based approaches
10	DM II	Gradient descent, MLE and odds	Gradient descent (including for linear regression), maximum likelihood estimation, odds and log-odds
11	DM II	Logistic regression	The logit function, likelihood, recovering probabilities from log-odds
12	DM II	Support vector machines	Hard-margin (primal / dual), soft margin, kernel mapping, multiclass strategies, advantages and limitations
13	DM II	Neural network	Artificial neurons and perceptrons, learning rules, activation functions, MLPs, backpropagation, training issues and regularization
14	DM II	Deep neural networks	Activation functions, training and backpropagation, objective functions, tips and tricks, CNNs and RNNs
15	DM II	Ensemble learning	Bagging, boosting (AdaBoost, stumps), random forests, gradient boosting (XGBoost, LightGBM, CatBoost, EBM)
16	DM II	Explainability	TREPAN, LIME, LORE, SHAP, integrated gradients, instance-based explanations
17	DM II	Transactional clustering	K-Modes, ROCK, CLOPE, TX-Means
18	DM II	Sequential pattern mining	Generalized Sequential Patterns (candidate generation, pruning, support counting), time constraints
19	DM II	Time series	Visualization, missing values, anomalies, normalization techniques, components, models, decomposition
20	DM II	Similarity	Structural-based and shape-based similarity measures (including DTW)
21	DM II	Time-series clustering and classification	Similarity-based clustering, motif and discord discovery (matrix profile), classification and shapelet extraction