INFOMDWR: Course Syllabus 2023

Authors

Erik-Jan van Kesteren

Hakim Qahtan

Ayoub Bagheri

Published

May 16, 2024

1 Introduction

Data do not fall from heaven, but are created, manipulated, transformed, and cleaned - in any data analysis, therefore, the treatment of the data itself is just as important as the modeling techniques applied to them. In this course, you will get acquainted with and implement a variety of techniques to go from raw data to analyses, visualizations and insights for science and business applications. This is an overview course designed to give you the tools and skills to use and evaluate data science methods.

The course consists of two parts, data wrangling and data analysis, which are intertwined.

Prerequisites

We assume that students who will join the course will have knowledge of statistics up to regression and analysis of variance, as well as some experience in programming in languages such as R and Python.

Objectives

At the end of this course, students have attained the following objectives:

  • Know, explain, and apply data retrieval from existing relational and nonrelational databases, including text, using queries built from primitives such as select, subset, and join both directly in, e.g., SQL and through the python & R programming languages.
  • Know, explain, and apply common data clean-up procedures, including missing data and the appropriate imputation methods and feature selection.
  • Know, explain, and apply methodology to properly set-up data analysis experiments, such as train, validate, and test and the bias/variance trade-off.
  • Know, explain, and apply supervised machine learning algorithms, both for classification and regression purposes as well as their related quality measures, such as AUC and Brier scores.
  • Know, explain, and apply non-supervised learning algorithms, such as clustering and (other) matrix factorization techniques that may or may not result in lower-dimensional data representations.
  • Be able to choose between the different techniques learned in the course and be able to explain why the chosen technique fits both the data and the research question best.

2 Course Policy

This course is worth 14 ECTS, which means it is designed to give a full-time workload.

Weekly course flow

A regular week in this course consists of three lectures (Monday-Wednesday morning) and three lab sessions (Monday-Wednesday afternoon). The material is introduced on a theoretical level in the lectures and then put into practice in the lab sessions. The practical work done in these labs is drawn from real life situations that allow the students to experience how to solve data science problems.

In addition, students will spend time during each week on bi-weekly take-home group assignments.

  • The lectures are in-person. The required readings should be read before the lecture. These are not optional.
  • The lab sessions are in-person interactive sessions in which you apply the methods you learn about in the lectures. The answers to the exercises in the labs are discussed at the end of each session.
  • The skills acquired in the lectures and the labs provide the basis for doing the bi-weekly take-home assignment. This assignment is made in groups of 3-5 students and handed in via Blackboard.

Synchronous course policy

  • INFOMDWR is an offline-first course, with mostly in-person lectures and lab sessions.
  • We find it important for interactive and collaborative learning that the course is offline-first.
  • Due to a lack of space at UU, some lectures will be hybrid with maximum of 50-60 in-person spots. You will receive a teams meeting invitation for these sessions so you can follow online if you prefer. Note: these lectures will still be synchronous, so no recordings will be made available as per university policy.
  • If you miss a session, e.g., due to sickness, you should catch up in the regular way:
    • Read the readings
    • Go through the lecture slides
    • Do the practicals
    • Ask your peers if you have questions
    • (after the above) ask the lab teacher for further explanation

Who to ask what

There are many teachers in this course. If you have questions, first ensure the answer isn’t in this syllabus and then follow the table below:

Question type How to ask
Course proceedings Email course coordinators
Content - general Email / ask lab teachers
Practical content Email / ask lab teachers
Assignment content Email / ask lab teachers
Lecture content Email Lecturer (Hakim Qahtan / Daniel Oberski)

Grading policy

Your final grade in the course consists of the following parts:

  • Biweekly assignments (20%): every two weeks, there is a group assignment. Each assignment is graded and worth 5%.
  • Midterm exam (40%): Halfway through the course, there is a midterm exam. The content of this exam is everything in the first half of the course.
  • Final exam (40%): At the end of the course, there is a final exam. The content of this exam is focussed on the second half of the course (but note that the material builds on the first half).

To pass the course:

  • The final grade should be greater than or equal to 5.5.

Resit:

  • If you obtain a grade between 4.0 and 5.4 on the midterm or the final exam, you are eligible for the resit. The grade attained in the resit will replace the grade from the exam.
  • You can only retake one of the two exams. By default this will be the one with the lowest grade (between 4.0 and 5.4). If you obtained the same grade for both, you will redo the final exam.

3 Course materials

Required Software

In this course, we will use a variety of software, but mainly SQLite, Python and R. Try to install both on your computer by the start of the course; we will also have a set-up computer lab on the first day to help you with this process.

Installing DB Browser for SQLite For the SQL parts, we recommend installing DB Browser for SQLite. Installation instructions for mac, windows, and linux can be found here.

Installing Python & Jupyter For the python parts of the course, we will use Google Colab, which is an interactive online notebook environment; this means no installation is necessary! However, you do need a google account, so make sure you have one (or make one specifically for the course).

Installing R & RStudio First, install the latest version of R for your system (see https://cran.r-project.org/). Then, install the latest (desktop open source) version of the RStudio integrated development environment (link). We will make extensive use of the tidyverse suite of packages, which can be installed from within R using the command install.packages("tidyverse").

Required readings

Freely available sections from the following books:

Book Title (Authors) URL
DBSC Database System Concepts (Silberschatz, Korth, Sudarshan) db-book.com
MMDS Mining Massive Datasets (Leskovec, Rajaraman, Ullman) mmds.org
PDSH Python Data Science Handbook (VanderPlas) jakevdp.github.io/PythonDataScienceHandbook
DMCT Data Mining: Concepts and Techniques (Han, Kamber, Pei) Find it online for free.
R4DS R for Data Science (Grolemund & Wickham) r4ds.hadley.nz
ISLR Introduction to Statistical Learning (James et al.) statlearning.com
DLBK Deep Learning (Goodfellow, Bengio, Courville) deeplearningbook.org
FIMD Flexible Imputation of Missing Data (van Buuren) stefvanbuuren.name/fimd
SLP3 Speech and language processing (Jurafsky & Martin) stanford.edu/~jurafsky/slp3
TTMR Text mining with R: A tidy approach (Silge & Robinson) tidytextmining.com/
OMNG Operations Management 4th ed. (Reid & Sanders) Find it online for free.

And (parts of) the following references:

  • Oberski, D.L. (2016). Mixture models: Latent profile and latent class analysis. Modern statistical methods for HCI, 275-287. URL
  • Mehrabi, N., Morstatter, F., Saxena, N., Lerman, K., & Galstyan, A. (2021). A survey on bias and fairness in machine learning. ACM computing surveys (CSUR), 54(6), 1-35. URL
  • Hennig, C. (2015). Clustering strategy and method selection. arXiv preprint arXiv:1503.02059. URL
  • Bouveyron, C., Celeux, G., Murphy, T. B., & Raftery, A. E. (2019). Model-based clustering and classification for data science: with applications in R (Vol. 50). Cambridge University Press.
  • Some other freely available articles & chapters

4 Class Schedule

You can find the up-to-date class schedule with locations on mytimetable.uu.nl.

Week Date Topic Type Reading
1 2023-09-07 Introduction to the course Lecture Syllabus
1 2023-09-07 Introduction lab: setting up your computer Lab PDSH 1 and R4DS 3, 5, 7
1 2023-09-08 Data models Lecture DBSC 1, 3.2, 3.9
1 2023-09-08 Data models Lab
2 2023-09-11 Data extraction with SQL Lecture DBSC 2.6, 3.3 - 3.8
2 2023-09-11 Data extraction with SQL Lab
2 2023-09-12 Integrity constraints in databases Lecture DBSC 3.2.1, 4.4, 6.1, 6.2
2 2023-09-12 Integrity constraints in databases Lab
2 2023-09-13 Functional Dependency & Data Integration Lecture DBSC 7.1-7.4.1
2 2023-09-13 Functional Dependency & Data Integration Lab
3 2023-09-18 Hetero. data analysis & string similarity Lecture MMDS 3.1-3.5
3 2023-09-18 Hetero. data analysis & string similarity Lab
3 2023-09-19 Data extraction in Python Lecture PDSH 3
3 2023-09-19 Data extraction in Python Lab
3 2023-09-20 Data preparation 1 Lecture DMCT 3.1, 3.2, 12.1 - 12.4
3 2023-09-20 Data preparation 1 Lab
4 2023-09-25 10:00AM assignment 1 Deadline
4 2023-09-25 Data preparation 2 Lecture DMCT 3.3-3.5
4 2023-09-25 Data preparation 2 Lab
4 2023-09-26 Guest lecture on cloud computing Lecture TBD
4 2023-09-26 Cloud computing Lab
4 2023-09-27 Data visualization Lecture R4DS 3, 5, 7 (optionally 2-8)
4 2023-09-27 Data visualization using ggplot Lab
5 2023-10-02 Exploratory data analysis Lecture R4DS 3, 5, 7 (optionally 2-8)
5 2023-10-02 Exploratory data analysis in R Lab
5 2023-10-03 Supervised learning: Regression Lecture ISLR 1, 2.1, 2.2, 3.1, 3.2, 3.3, 3.5
5 2023-10-03 Supervised learning: Regression models in R Lab
5 2023-10-04 Q&A Lecture
5 2023-10-04 No lab, time to study Lab
5 2023-10-06 Midterm exam Exam
6 2023-10-09 10:00AM assignment 2 Deadline
6 2023-10-09 Supervised learning: model evaluation Lecture ISLR 5.1, 8.1, 8.2.1, 8.2.2, 8.2.3
6 2023-10-09 Supervised learning: model evaluation Lab
6 2023-10-10 Supervised learning: classification Lecture ISLR 4.1, 4.2, 4.3, 4.4.1, 4.4.2
6 2023-10-10 Supervised learning: classification Lab
6 2023-10-11 Deep learning Lecture ISLR 10, DLBK 11, (optionally 6)
6 2023-10-11 Deep learning Lab
7 2023-10-16 Missing data 1: Mechanisms Lecture FIMD 1.1, 1.2, 1.3, 1.4
7 2023-10-16 Missing data mechanisms Lab
7 2023-10-17 Missing data 2: Solutions Lecture FIMD 1.1, 1.2, 1.3, 1.4
7 2023-10-17 Imputation methods Lab
7 2023-10-18 Clustering Lecture ISLR 12.1, 12.4
7 2023-10-18 Clustering Lab
8 2023-10-23 10:00AM assignment 3 Deadline
8 2023-10-23 Model-based clustering Lecture Oberski (2016), optional Hennig (2016), Bouveyron et al. (2019)
8 2023-10-23 Model-based clustering using MClust Lab
8 2023-10-24 Text mining 1 Lecture SLP3 2.1, 2.4, 6.2, 6.3, 6.5, 6.8; TTMR 3
8 2023-10-24 Text mining 1 Lab
8 2023-10-25 Text mining 2 Lecture SLP3 2.1, 2.4, 6.2, 6.3, 6.5, 6.8; TTMR 3
8 2023-10-25 Text mining 2 Lab
9 2023-10-30 Time series Lecture OMNG pages 265-294 (forecasting)
9 2023-10-30 Time series Lab
9 2023-10-31 Data streams Lecture MMDS 4.1 – 4.4
9 2023-10-31 Data streams Lab
9 2023-11-01 Algorithmic fairness Lecture Mehrabi et al. (2021)
9 2023-11-01 Algorithmic fairness Lab
10 2023-11-06 10:00AM assignment 4 Deadline
10 2023-11-07 No lecture: study time Lecture
10 2023-11-07 No lab: study time Lab
10 2023-11-08 Q&A Lecture
10 2023-11-08 No lab: study time Lab
10 2023-11-10 Final exam Exam
10 2024-01-15 Resit exam Exam