Искусственный интеллект в науках о здоровье и социальных науках (курс на английском языке)
Факультет государственного управления

This course suggests approaches to understanding the usage and implementation of the artificial intelligence solutions in the social and health sciences. The course is particularly focused on the important concepts of interpretability, fairness, and generalizability in the health and social sciences. Machine learning is widely used for descriptive tasks in the sciences, aiming at a quantitative summary provision of certain features of the world. Needless to say, description proves to be the basis of all applied research as far as it is essential to quantify the phenomenon under study. In respect to the social sciences we need to assess prevalence or distributions of variables between social groups, countries and we often need to estimate it over time. The will provide approaches to pre-selection of variables for the factor analysis that can be considered interpretable and meaningful. We will discuss the role and ML approaches to risk profiles identification, will focus on the data that originate from sensor data and can be used to detect conditions. Further to this task it is highly important to understand the limitations of the approach. One of the most serious issues to discuss while implementing technologies the that of responsibility and trust. By illustrations in health-related field we are going to find out peculiarities of medical decisions, evaluating care and ensuring further interaction. Globally, strategy on digital health development relies greatly on new technologies implementation in reducing costs, preserving the quality of care and overcoming inequalities, ensuring better health and patient outcomes. At the same time people are concerned about technologies that might take their jobs and lead to the skill atrophy. Artificial intelligence proves to be one of the most promising domains worldwide. However, the hype around the concept provided the avalanche of potential technical solutions, however, without proper attention to the stakeholder analysis, ethics and bioethics involved. Also, the background of the technologies that have been used deserves attention. We are going to contemplate what emotions can be interpreted by AI and how. We will analyze the potential benefits of AI technologies and pay attention to the unintended consequences.