PENG9900 Doctoral Thesis in Engineering Science Course description

The thesis is an independent scientific work that shall meet international standards in terms of ethical requirements, academic level, and methodology. The thesis shall contribute to developing new academic knowledge in engineering science, and may be submitted either as a compilation comprising a minimum of three articles or as a monograph.

Conditions for submitting the thesis for assessment:

The training component of the programme must be approved before students may submit their thesis.

This course provides a broad introduction to machine learning (ML), which includes supervised, unsupervised, and reinforcement learning, and deep learning (DL) that can be used in different application domains. Students will learn both theories and practices in ML and DL. Moreover, students will learn from studying, presenting, and discussing relevant research articles and expose themselves to research by doing a research project.

Teaching during work on the thesis is provided in the form of supervision. Working methods for writing the thesis consist mostly of self-study and research activities, as well as participation in research communities, presentations of the student's own research in research fora, research communities, and at international scientific conferences.

All PhD students must present their projects at interdisciplinary seminars at the start, midway, and near the end of the programme. The midway presentation should ideally be conducted in English and be followed by a discussion on current progress between the PhD student, the supervisor(s), and the PhD Programme Director.

Students are required to present their research at at least one international scientific conference.

On successful completion of the course, students should have the following learning outcomes defined in terms of knowledge, skills, and general competence.

Knowledge

The student has:

• knowledge of supervised, unsupervised, reinforcement learning
• good understanding of the principles of state-of-the-art deep neural networks such as convolutional neural networks, sequential models (RNN, LSTM), Transformers, GenerativeAI (Autoencoder, GAN, Diffusion models), and reinforcement learning.
• a good understanding of both theoretical and practical know-how required to use machine learning and deep learning methods effectively.

Skills

The student can:

• build, train, test, and deploy machine learning and deep learning models
• analyze machine learning methods in regard to their performance and effectiveness
• use existing deep learning networks, improve and/or customize them to apply to new problems

General competence

The student:

• has both theoretical and practical understanding of machine learning and deep learning methods
• can discuss relevance, strength, and limitations of machine learning and deep learning in solving real-world problems
• can work on effectively relevant research projects

The course consists of lectures, group consultations, presentation seminars, and project work. In the seminars, students will read papers, present, and also actively participate in other presentations. This will facilitate research-oriented education in the field. Research projects will be aimed at cultivating the students towards good future researchers.

The following required coursework must be approved before the student can take the exam:

• Two oral presentations (one on a given topic, one on the topic of own choice)
• Participate as a prepared opponent/discussant in two presentations from other students

There is mandatory attendance in obligatory consultation meetings and a minimum of 80% mandatory attendance in the lectures.

Students who do not meet this requirement will not be allowed to sit the exam.

Exam in two parts:

• A group project: implementation and report (about 7000 words). A group of 2-3 students will be formed during the course. Each group member receives an individual grade based on their contribution to the project.
• Individual oral exam (about 30 minutes).

Each of them carries 50% weight in the final grade. The oral examination cannot be appealed.

Both exams must be passed in order to pass the course.

New/postponed exam

In case of failed exam or legal absence, the student may apply for a new or postponed exam. New or postponed exams are offered within a reasonable time span following the regular exam. The student is responsible for registering for a new/postponed exam within the time limits set by OsloMet. The Regulations for new or postponed examinations are available in Regulations relating to studies and examinations at OsloMet.