Curriculum

In addition to the security and privacy of our everyday life, uses of cryptography have been continuously expanding from quantum cryptography to blockchain/cryptocurrency. Instead of understanding detailed mathematical theories behind cryptography, the purpose of this class is to learn basic cryptography, cryptographic protocols, and the current and future applications of cryptography. As a case study, we will review details of the blockchains technology.

This course introduces the fundamentals of reinforcement learning in a way that is accessible to undergraduate students. It covers the minimum required mathematics for understanding reinforcement learning and helps students develop interest through simple examples and Python-based exercises. The course covers classical reinforcement learning topics such as Markov Decision Processes, dynamic programming, TD learning, and Q-learning, as well as recent advances in deep reinforcement learning.

This course will survey key concepts of quantum mechanics, solid-state physics, and semiconductor physics in view of realizing nano/quantum electronic devices. Rather than the traditional approach to semiconductor devices based on the drift-diffusion equation, the first-principles approach starting from quantum transport theory will be presented.

This course covers machine learning techniques to analyze visual data. Specifically, this course focuses on fundamental machine learning and recent deep learning methods that are widely used in visual data analysis, and discusses how these methods are applied to solve various problems with visual data. This course consists of lectures, practices, and projects.

System software is a fundamental driving force that lets the applications to interact with the computer hardware. The students will learn the role, the internal design and implementation of the system software including shell, linking, loading and operating system internals.As a reference operating system, we will use xv6 and Linux.

This course covers variety of audio processing techniques for Virtual Reality (VR), 3D Audio, Room Impulse Responses, Basic Filter Design, and Sound Source Localization. Basic principles of sound propagation and human hearing are explained with listening examples. Applications and exemplary implementation of individual topics are presented with Matlab codes. Single channel filtering, time-frequency analysis, multichannel signal processing are major tools utilized for these applications. This course also offers term projects in which students can experience one of these techniques by their own. The course is designed to practice the knowledge learned from Signals and Systems & Digital Signal Processing.
Main text: lecture slides, Prerequisites: EE202 Signals & Systems

The objective of this course is to cultivate the practical ability to design AI models through the deep
understanding of various DNN models and hands-on experiments. This course will be a project-based
and experiment-oriented class in which students form a team with a topic selected from AI-related
challenges. Team members can select the topic from two major themes (e.g., anomaly detection,
reinforcement learning) proposed by lecturers.

The objective of this course is to cultivate the practical ability to design AI models through the deep understanding of various DNN models and hands-on experiments. This course will be a project-based and experiment-oriented class in which students form a team with a topic selected from AI-related challenges. Team members can select the topic from two major themes (e.g., anomaly detection, reinforcement learning) proposed by lecturers.

We discuss many philosophical issues arising in artificial intelligence (AI) research. We will try to answer many fundamental questions such as whether human-like AI is possible, nature of consciousness, self-awareness, qualia, free will, whether the law of physics allows non-biological human-like AI, and moral issues with AI.

The theme of this course is to identify and use patterns in intelligent autonomous systems. By studying patterns, many control and state-estimation algorithms can be more efficient in data consumption and computation time. Relevant motivating applications include robotic path-planning with AI, fault-tolerant control, and decision-making networks (e.g., vehicle traffic systems, UAV traffic management).