COURSE DESCRIPTION:
This elective course will provide an introduction to modern methods in Artificial Intelligence (AI), particularly focusing on AI for biomedical research, through frontal classes and critical reading and discussion of selected papers.

During the last decade, AI has achieved impressive progress in real-world applications that appeared out of reach just a few years ago, and now features regularly in both scientific and general news. Modern AI methods are also becoming increasingly recognized as an integral component of the toolbox of the biomedical research community. The goal of this course is to provide an introduction to the most important paradigms in modern AI, through lectures, critical reading and discussion of research papers relevant for graduate students in biomedical sciences. Each of the selected papers adopts those algorithms in one of three ways: 1) as a powerful tool for analysis of complex and/or large scale data (e.g. “alpha-fold”, a deep-learning algorithm for protein folding); 2) as a core component in semi-automated medical applications (e.g. the “UNet” for analysis of CT scans and x-rays); 3) as a computational model of biological processes (e.g. “deep convolutional neural networks” to explain the logic of neural activity in visual areas of the brain).

The course will last 7 weeks, with a total of 7 frontal classes and 7 paper discussion classes. The contents of each class are detailed below in the list of lectures.

COURSE OBJECTIVES:
The objective of the course is to learn foundational techniques in machine learning, how they are used in modern AI methods, and their application to biomedical research, through a combination of frontal lectures and group discussion of selected papers.

After the course, students will have acquired literacy and experience with relevant applications of AI. This will enable them to communicate fluently with AI experts, both in academia and industry, as well as provide the starting point for those who may be interested in applying modern AI methods in their thesis research.

PREREQUISITES:
Recommended course: Introduction to the Mathematics of Theoretical Systems Biology (Block I).

It is recommended that students be familiar with the following maths prior to enrolling in this course:

• Linear algebra (matrix-vector operations; tensors).
• Calculus (integrals, derivatives and function optimization).
• Probability and statistics (Marginal/conditional/joint distribution; Bayes rule).

Students can use this online resource https://compneuro.neuromatch.io/tutorials/intro.html in the ‘pre-reqs refresher’ section for introductory tutorials.

REQUIRED MATERIALS:
Reading materials will be provided for each class.

SUITABLE FOR 1ST YEAR STUDENTS:
Yes.

STUDENT ASSESSMENT:
Two or three papers (depending on class size) will be discussed in each discussion class. Each paper is presented by one group of students, and criticized by one other group. There will be a 15-20 minute presentation, followed by 15-20 minute Q&A/scientific debate between the two groups.

Grades will be based on the understanding of papers, demonstrated by the clarity of presentation of the assigned papers and by the relevance of questions to other presenters. Specifically, when presenting a paper, students need to clearly explain the scientific background (in the context of related scientific literature), the main hypothesis or challenges addressed by the paper, the methods used to perform the work, whether the results presented are sound and compelling, and the limitations and broader implications of the paper. When criticizing the paper presented by the other group, students will be assessed on whether the questions/criticisms are relevant (e.g. they address a conceptual or methodological issue of the paper, or they highlight important limitations of the paper) and on whether they support their criticism with pointers to other literature or with specific pointers to the text/figures of the paper.

Although students present and criticize papers as a group, each student is expected to present a section of their paper, and to ask questions/criticism about the other group’s paper; therefore, each student will receive an individual grade.

In addition, class participation will be evaluated for frontal lectures: specifically, whether students ask relevant questions, and answer when questions are posed by the lecturer. Paper discussion grades account for 80% of final grade, class participation 20%.

The final grade (H/P/F) is cumulative and will be transmitted at the end of the course. A grade of H can be obtained by consistently (i.e. throughout all classes) showing knowledge and understanding beyond the assigned papers, for instance, by identifying other papers related to the assigned paper and finding deep connections or conflicts between them. A student will receive F if they consistently fail to present clearly their paper, or to ask relevant questions about the other papers.

COURSE ATTENDANCE POLICY:
Students are expected to attend all classes. If missing a discussion class, students will write a 1 to 2 page (plus figures and bibliography, optionally) critical review of the two papers.

The final grade (honors/pass/fail) is cumulative and will be transmitted at the end of the course. A grade of honors can be obtained by consistently (i.e. throughout all classes) showing knowledge and understanding beyond the assigned paper/tutorial, for instance, by identifying other papers related to the assigned paper and finding deep connections or conflicts between them. A student will receive a fail if they consistently fail to present clearly their paper, or to ask relevant questions about the other papers.

CREDIT HOURS: 2.0