Guideline for Students writing a Bachelor or Master’s Thesis
in the Intelligent Systems and Machine Learning (ISML) Group
General procedure, steps, and milestones Writing a thesis in our group will involve the following main activities and milestones (to be explained in further detail below):
Finding a thesis topic: Students can choose from our list of open topics or make own suggestions. The time required for finding a topic depends on the student’s efforts and her/his familiarity with the research of the ISML group.
Writing and submitting the thesis proposal: Writing a thesis proposal is important to specify the topic and define the goals of the thesis. This activity includes familiarization with the topic and normally requires about 2–4 weeks. The time spent for the proposal is not considered as part of the thesis. You will be registered officially upon successful submission of the proposal.
Starter talk: Shortly after the start of the thesis, it is common to give a short presentation in the seminar of the research group. The talk is meant to explain the topic and the goals of the thesis, and to sketch some first ideas of how to tackle these goals, so that the members of the group can give comments and provide feedback.
Working on the thesis: In order to accomplish and successfully complete your thesis, you need to organize your time between reading relevant literature and related work, solving your own tasks, developing and implementing your solutions, and writing the thesis. As a research group, we offer you support through regular meetings with the supervisor, our students seminar, a suitable working environment, and proofreading before the official submission of the thesis.
Submitting the thesis: At this stage, you are required to submit hard copies of the written thesis, supplemented with your source code, experimental results, a PDF of the thesis, etc., in electronic from.
Oral defense: The defense will take place after the formal submission. Defending the thesis prior to the formal submission is only possible under specific circumstances and in agreement with the supervisors.
Prerequisites Students interested in writing a thesis in our group should have a background in our main research topics. Normally, we assume that you attended (and successfully passed) some of our lectures (Foundations of Intelligent Systems or Data Mining in the case of Bachelor students, and Machine Learning I + II in the case of Master students). Exceptions to this rule are possible but need to be justified.
Topics A list of open topics for Bachelor and Master’s theses can be found on the ISML website. Please have a look at this list prior to contacting Prof. Hüllermeier or any other group member (contact persons are also indicated on each thesis proposal). Ideally, you preselect two or three topics according to your preferences, and then ask for an appointment to discuss about details. Needless to say, you are also invited to propose topics yourself. In this case, please prepare a one-page summary of your main ideas.
Supervision and co-supervision Prof. Hüllermeier will be the main supervisor for all theses written in the ISML group. Additionally, you will be assigned one of the research assistants as a co-supervisor and primary contact person. Please make sure that you meet and/or communicate with the co-supervisor on a regular basis, so as to keep her or him informed about the progress as well as any kind of difficulties that may arise. As a rough guideline, you should get in touch with co-supervisor at least every two weeks. Students are encouraged to organize their work and share it with the supervisor using the university’s GitLab service; in this way, sending emails with large attachments can also be avoided.
Proposal Once the topic has been fixed, a proposal for the thesis needs to be prepared. This proposal is a prerequisite for registering the thesis. It is supposed to provide a motivation of the topic, to elaborate on the state of the art and prior work, and to explicate the goals of the thesis. The goals should be defined carefully, as they provide the basis for the final assessment of the thesis. The proposal must also contain a preliminary structure of the thesis and a schedule (Gantt chart). The preparation of an acceptable proposal—with appropriate assistance and in due course—is a first contribution and a main prerequisite for registration of the thesis. We reserve the right to withdraw our commitment of supervision in cases where this prerequisite is not met.
Template The thesis should be written in LaTex and adhere to our template.
Proofreading The co-supervisor will normally offer you to proofread a draft of your thesis, either as a whole or in parts (e.g., single chapters). The goal of proofreading is to provide general feedback and give hints for possible improvements (but not to make detailed corrections). Please note that proofreading will be offered only once for each part of the thesis; therefore, you should make sure that your draft is sufficiently mature. Moreover, please take into consideration that feedback cannot be expected on a very short notice; typically, it will take about two weeks.
Working environment You are kindly invited to work in the PC pool of the ISML group, unless space restrictions prevent us from offering you a place equipped with a computer. Of course, if you prefer, you can also work at home or any other place. In special cases, we can support you in acquiring a virtual machine from the IMT, or to get access to HPC resources (including GPUs).
Oral defense The thesis will end with an oral defence in the form of a seminar. The seminar will start with a presentation of the topic and main results of the thesis, followed by an informal discussion in which the audience can ask questions. The presentation should not exceed 20 minutes in the case of Bachelor and 25 minutes in the case of Master’s theses. There is no time limit for discussion.
Evaluation of the thesis The final assessment of the thesis will be based on the following criteria:
Working attitude Familiarization with the topic; independency, autonomy; commitment and engagement; utilization of resources (tools, computer infrastructure, etc.).
Contents and results Documentation of related work; solution of the tasks and compliance with objectives; quality of the solutions; documentation und reproducibility.
Written thesis Structure and readability of the thesis; length of the thesis, balance between breadth and depth; formal presentation and correctness; linguistic quality (grammar, orthography, typos, etc.); extent and quality of illustrations, figures, tables, etc.; citation and list of references (coverage, correctness, etc.).
Defense of the thesis Quality of presentation; quality of slides; response to questions; compliance with time limit.