IST 4520 · Fall 2026

Data Mining

Credits: 3
Lecture: 18 h
Lab: 12 h
Self-study: 90 h
Type: Elective

This week

Week 1 Introduction to Data Mining & CRISP-DM

Welcome to IST 4520 — Data Mining. This session introduces the course structure, the four types of data mining problems, and the CRISP-DM process that frames every project we will build this term.

Read North (2012), Chapter 1 — Introduction to Data Mining and CRISP-DM
Install Orange Data Mining before Week 2 (instructions in Lab 0) due Before Week 2
Join the course channel and register on Kaggle

A hands-on introduction to data mining for business decision-making. The course equips students with business-analytic thinking: framing the right question, selecting the right technique, and interpreting results in actionable business language — not writing algorithms from scratch.

Students work through the full CRISP-DM lifecycle using Orange Data Mining (a no-code, drag-and-drop platform) supplemented by light Python/pandas scripts. Core techniques covered include data exploration and visualization, association rules, k-Means clustering, KNN and Naïve Bayes classification, and linear and logistic regression. Midterm assessments include a practical exam on Orange and a group project presentation; the final is a written exam.

Full syllabus Labs & assignments Resources

Schedule

Fifteen weeks alternating lecture and lab. Lab weeks include a business memo and, where noted, a group progress milestone. Weeks marked TBD are not yet finalised and their materials are not yet published.

Week	Type	Topic	Materials	Assessment
1	Lecture	Introduction to Data Mining & CRISP-DM What is data mining and why it matters for business decisions Four problem types: Describe, Relate, Group, Predict Data concepts: rows, columns, variable types, structured vs unstructured The CRISP-DM six-stage process — the backbone of the course Orange Data Mining overview and live demo Quick pandas review: DataFrames, filtering, and descriptive statistics	Slides	—
2	Lab	Lab 0 — Environment Setup & First Orange Project Install VS Code + Python (or set up Google Colab) Install Orange Data Mining First Orange workflow: exploring the HDI economic dataset Mini-memo: turning scatter plot findings into a business insight	—	Lab 0 submission
3	Lecture	Understanding & Preparing Data TBD Variable types and roles (target, feature, meta) Data warehouse, data mart, and data lake concepts Data quality: missing values, outliers, duplicates Normalisation, standardisation, and creating new variables	—	—
4	Lab	Lab 1 — pandas Introduction & Data Cleaning TBD pandas: reading data, selecting and filtering, descriptive stats, handling missing values Data cleaning in Orange: Impute, Distributions, Scatter Plot, Box Plot Business memo: documenting data quality findings	—	Lab 1 + Excel exercise
5	Lecture	Finding Relationships — Correlation & Association Rules TBD Pearson correlation coefficient and the correlation matrix Association rules: support, confidence, and lift Market-basket analysis and product recommendation applications	—	—
6	Lab	Lab 2 — Correlation & Market-Basket Analysis TBD Correlation matrix and scatter plots in Orange Association rules with the Orange Frequent Itemsets widget Business memo: interpreting association rules for a retail context Current-event Presentation #1	—	Lab 2 + Presentation #1
7	Lecture	Clustering — k-Means & Customer Segmentation TBD k-Means clustering: intuition and algorithm Choosing the number of clusters k (Silhouette score) Customer segmentation application Midterm Exam 1 review: CRISP-DM, data preparation, association, clustering	—	—
8	Assessment	Midterm Exam 1 — Practical Exam on Orange 90-minute individual practical exam Tasks: load a provided dataset → preprocess → build a model → write a business interpretation Covers Weeks 1–7: CRISP-DM, data preparation, exploration, association, clustering	—	Midterm Exam — Part 1 (20%)
9	Lecture	Classification I — KNN & Naïve Bayes TBD Classification problem setup: train/test split, the confusion matrix k-Nearest Neighbours (KNN) Naïve Bayes classifier Evaluation metrics: accuracy, precision, recall, F1-score	—	—
10	Lab	Lab 3 — Classification in Orange TBD Test & Score and Confusion Matrix widgets in Orange ROC curves and model comparison Business memo: which model works best and why it matters	—	Lab 3
11	Lecture	Prediction — Linear & Logistic Regression TBD Linear regression: fitting a line, R², and residuals Logistic regression for yes/no business outcomes Interpreting coefficients in business language	—	—
12	Lab	Lab 4 — Regression & Group Project Launch TBD Regression in Orange and with Python pandas Reading and adapting a Python regression script Group project: confirm team, topic, and dataset	—	Lab 4 + Confirm group project topic
13	Lecture	Advanced Topics — Decision Trees, Neural Networks & Text Mining TBD Decision Trees and the Orange Tree Viewer Neural networks: intuition and business applications Text mining and NLP overview Cross-validation and trustworthy model evaluation	—	—
14	Assessment	Midterm Exam 2 — Group Project Presentations TBD Group presentations: 10–12 minutes per team + Q&A Current-event Presentation #2 Course review and final exam preparation	—	Midterm Exam — Part 2 (20%) + Presentation #2
15	Assessment	Final Exam TBD 90-minute written exam Covers the full course: CRISP-DM, all techniques, model selection and interpretation Question types: scenario-based technique selection, reading model output, short essays	—	Final Exam — 50%

Teaching staff

Dr. Trong-Nghia Nguyen

Course Instructor

nghiant@neu.edu.vn

Room 1613, Building A1

Office hours: By appointment

Personal website

MSc. Le Duy Khanh

Teaching Assistant

khanhld@neu.edu.vn

Lab troubleshooting and homework help

Faculty profile