PCS956 Research Trends in Applied Machine Learning

Contents and structure

Machine learning is an area within computer science where computer systems are designed to learn from large sets of examples. Machine Learning has recently gained great importance in domains such as natural language processing, image recognition, recommendation systems, the design of search engines, robots, and sensor systems, for extracting information from large scientific data sets. The focus of this course is on emerging trends in applied machine learning. Application areas discussed in the course include topics from computer vision, robotics, energy, and computational medicine and biology. After a joint common introduction, the students in the course will select a course track module focusing on a specific application domain. The module includes deep learning for computer vision: introduction to methods, software frameworks and a selection of applications; bayesian networks and causal inference; optimization in knowledge recommendation and privacy/security. The course modules may be adapted depending on the research interest of the participants and lecturers.

Learning Outcome

A student after the course the candidate should be able to: Knowledge

• explain the mathematical methods underlying machine learning, as well as how these methods are applied in selected application domains.
• assess machine learning problems in order to select and apply suitable methods and software tools to solve them.
• formulate and approach a new machine learning problem setting in a creative, critical, and systematic way.

Skills

• design solution strategies for machine learning problems,
• implement machine learning based solutions for one or more of the application domains covered in the course.

General competence:

• develop solutions for real-life problems,
• work in multidisciplinary environment,
• technical communication.

Entry requirements

General admission criteria for the PhD programme.

Recommended previous knowledge

A solid background in linear algebra, statistics, and familiarity with Python programming is recommended.

Teaching methods

The course consists of three core learning elements: Lectures on selected machine learning topics; seminars and workshops aimed at adopting and understanding a machine learning algorithm from recent literature (a prominent journal or conference in the field of machine learning), and then apply it on a selected use case; project at the end of the course.

Each section of the course will start with a set of lectures. The seminars/lectures in the course are student centric. The majority of in-class activities are comprised of students presenting their work to each other and receiving feedback.

Compulsory learning activities

Student activity expected during lectures/seminars, see teaching methods

Assessment

The course is graded pass/fail based on a project report/research paper and an oral exam. Each of the two components must result in a pass grade in order to obtain a pass grade for the entire course.

Acknowledgment: 

This course is developed by a partial support from RCN-INTPART DTRF Project.