AI for Good

In this seminar, students will apply AI methods on real world data to solve societal problems. Participants will analyse the data and apply different AI methods to visualize and gain insights on their data. Students are encouraged to work in teams of four. Topics are selected with a focus on having a positive impact on society. Students will summarize all findings in a report and a presentation.

  • DE - Bitte beachten: Es gibt keine deutsche Kursseite, da der Kurs nur in Englisch angeboten wird.
  • EN - Please note: This page has no German version because the course is only offered in English.

Quick Info: Master | Seminar | English | Summer Semester | 6 ECTS

Summary

CourseAI for Good
ChairInstitute of AI in Management
Lecturer Prof. Dr. Stefan Feuerriegel
AssistantsAnnually changing
Weekly HoursCoaching sessions by individual arrangement
Target GroupMMT & Master programs at the faculty of mathematics, informatics and statistics
ExaminationSeminar paper and video presentation, group work
PrerequisitesSufficient programming skills required (e.g., R, Python), such as taught in one of our previous courses (e.g., AI for Managers, Introduction to AI, etc.)
Course MaterialCourse material will be shared via Moodle, students are required to self-enrol to the course through Moodle, self-enrolment key can be accessed via LSF
RotationAnnual rotation, summer semester
LanguageEnglish
ECTS6

Description

  • This is our flagship research seminar. Our objective is to do a proper, rigorous, and novel research project end-to-end with you and where we guide you through all steps. The output should be an academic paper that is submitted to one of the top conferences in AI for Good, such as ICWSM or ACM.
  • We plan for extensive and detailed feedback sessions. This will be needed to ensure that we ramp you up to the standards of a scientific conference. For this, we will invite groups to additional meetings to have frequent project updates and provide extensive feedback (e.g., every 2-4 weeks). We will aim for Zoom at an evening slot, to avoid overlap.
  • We expect students to allocate a substantial amount of time, as we also do the same on our side. This seminar is thus best suited for those eager to really advance their skillset further, plan for an academic career, etc. - from experience, the paper re-writing process will need 2-3 cycles alone. If you prefer to have a more "traditional" seminar it may be better for you to choose one of our other seminars.
  • Grading depends on whether the objective of having a paper that is in publishable form at one of the above mentioned scientific outlets and thus of the corresponding scientific contribution has been achieved (i.e., where one would expect a positive response in the peer review process). We reserve very good and good grades exclusively for that. However, this should be doable if you invest sufficient effort.
  • By participating in the seminar, you agree that we submit the result as a paper and make the materials public. We do so by acknowledging your contribution in a fair manner according to good scientific practices. The intention is listing you as a co-author. However, if things don't work out well or if there is a skew in the contributions within your group (e.g., somebody is taking it "too easy"), we may retain the right to list someone in the Acknowledgements only.

We have carefully designed research questions that are novel and of direct scientific and public significance.

Outline

  • Final dates and times will be shared via Moodle.
  • All topics will be published on Moodle approximately one week before the semester starts.
  • The first session usually takes place in the first week of the semester, via Zoom.
  • It is important to participate in the first session, because groups will be set-up and topics will be distributed.
  • Coaching sessions by individual arrangement.
  • Submission of seminar paper and video presentation via Moodle, usually towards the end of the semester.

Literature

Literature depends on the topics.