DAT405: Introduction to Data Science and AI, lp1 HT22 (7.5 hp)
This page is under construction
The course is divided into three parts:
- Part I: Introduction to data science (3 weeks)
- Part II: Statistical methods in data science and AI (2 weeks)
- Part III: Introduction to AI (3 weeks)
Each part is in turn divided into weekly modules, with weekly assignments. The assignments are performed in student pairs, and the deadline for each assignment is Monday (23:59) the following week.
|1||Introduction to Data Science. Getting started with Python|
|2||Regression and classification|
|4||Bayesian statistics and graphical models|
|5||Markov models, kernel methods and decision trees|
|6||Introduction to AI and its ethics|
|7||Machine learning and neural networks|
Zoom links to each session can be found under Home.
Lecturer part I (week 1-3): Johan Brown-Cohen (firstname.lastname@example.org) Computer Science and Engineering
Lecturer part II (week 4-5): Marina Axelson-Fisk (email@example.com), Mathematical Sciences
Lecturer part III (week 6-8): Bastiaan Bruinsma (firstname.lastname@example.org), Computer Science and Engineering
- Adam Breitholtz
- Firooz Shahriari Mehr
- Lena Stempfle
- Mehrdad Faharani
- Linus Aronsson
Contact details are on the course Home page
The course has no mandatory textbooks, but the following books are recommended for consultation. The page for each module contains links to lecture slides and other supplementary material, including some links to video material.
- S.S. Skiena, The Data Science Design Manual, Springer, 2017. The Ebook (Links to an external site.) is available for free from the Chalmers library from within the Chalmers network:
- Jake VanderPlas. Python Data Science Handbook (Links to an external site.) , O’Reilly Media, Inc., 2016. (Links to an external site.)
- Jake VanderPlas. A Whirlwind Tour of Python (Links to an external site.) , O’Reilly Media, Inc., 2016.
- Allen B. Downey, Think Python: (Links to an external site.) How to Think Like a Computer Scientist (Links to an external site.) , 2nd edition . Green Tree Press, 2015. (Links to an external site.) (Links to an external site.)
Statistical methods for data science and AI
- [Bi]: M. Bishop, Pattern Recognition and Machine Learning (Links to an external site.) , Springer, 2006.
- [Ba]: D. Barber, Bayesian Reasoning and Machine Learning, Cambridge University Press, NY, 2012. (Links to an external site.)
- [M]: K. Murphy, Machine Learning: A Probabilistic Perspective, The MIT Press, 2012.
Changes made since the last occasion
No substantial changes in content.
Learning objectives and syllabus
- describe fundamental types of problems and main approaches in data science and AI
- give examples of data science and AI applications from different contexts
- give examples of how stochastic models and machine learning (ML) are applied in data science and AI
- explain basic concepts in classical AI, and the relationship between logical and data driven, ML-based approaches within AI.
- briefly explain the historical development of AI, what is possible today and discuss possible future development.
- use appropriate programming libraries and techniques to implement basic transformations, visualizations and analyses of example data
- identify appropriate types of analysis problems for some concrete data science applications
- implement some types of stochastic models and apply them in data science and AI applications
- implement and/or use AI-tools for search, planning and problem solving
- apply simple machine learning methods implemented in a standard library
- justify which type of statistical method is applicable for the most common types of experiments in data science applications
- discuss advantages and drawbacks of different types of approaches and models within data science and AI.
- reflect on inherent limitations of data science methods and how the misuse of statistical techniques can lead to dubious conclusions
- critically analyze and discuss data science and AI applications with respect to ethics, privacy and societal impact
- show a reflective attitude in all learning
The examination is through weekly assignments, carried out in student pairs.
The deadline for each week's assignment can be found at the foot of this page. You will have approximately 1 week from the last lecture on the topic until the deadline.
- Assignment grades
- Each assignment will be graded with 1-10 points
- Special rules apply for grades on late submissions without a valid reason (contact Adam):
- late <= 1 day (24h), can get max 9 points.
- 1 < late <= 2 days, can get max 8 points
- 2 < late <= 3 days, can get max 7 points
- 3 < late <= 4 days, can get max 6 points
- late > 4 days, can get max 5 points
- Course grades:
- Chalmers grades: Fail, 3, 4, 5
- The course grade will be based on the sum of the assignment grades
- At least 5 points on all Assignments are needed for a passing grade
- If the score is lower than 5 points you will have to resubmit, but the score on the resubmission cannot be higher than 5 points.
- No resubmissions are allowed unless under the above criteria.
- Point needed for final grade (must also pass all assignments):
- 5: >= 68
- 4: >= 56
- 3: >=40
There will be 8 assignments.
Grading will be based on a qualitative assessment of each assignment. It is important to:
- Present clear arguments
- Present the results in a pedagogical way
- Should it be table/plot? What kind of plot? Is everything clear and easy to understand?
- Show understanding of the topics
- Give correct solutions.
- Make sure that the code is well commented.
- Important parts of the code should be included in the running text and the full code uploaded to Canvas.
To obtain grade 3 or higher you must pass all the assignments.
The final grade is based on an overall assessment at the end of the course.
The syllabus page shows a table-oriented view of course schedule and basics of course grading. You can add any other comments, notes or thoughts you have about the course structure, course policies or anything else.
To add some comments, click the 'Edit' link at the top.