Special Topics

The course will cover the application of machine learning in CPU design, covering topics such as feature engineering, perceptrons, neural networks, recurrent neural networks, LSTMs, language models, generative adversarial learning, genetic algorithms, AI Agents, ML interpretation, reinforcement learning and other machine learning concepts for solving design challenges related to performance, power, and security; solving challenges common in many engineering fields. You will also get hands-on experience in training and inference of the latest Large Language Models on custom data. 

This course will introduce fundamental concepts in robot motion planning with a focus on spatial manipulators utilizing simulation and, if available, real robots. The course’s topics will include rigid-body spatial transformations, robot kinematics, trajectory generation, configuration space, and sampling-based path planning. Course projects and exercises will utilize a high-level programming language and modern tools and environments used in robotics.

This course focuses on applying advanced electromagnetic principles and semiconductor theory to design silicon photonic integrated circuits. Key principles such as matrix optics, waveguide theory, coupled mode theory, and P-I-N junctions will be used to design practical silicon photonic devices which are relevant in today’s foundries. Topics include passive wavelength filters, active switches, high-speed optical modulators, and photodetectors for optical communication and computing systems.

This course covers the fundamentals in algorithm design and analysis, focusing on the themes of efficient algorithms and intractable problems.  Topics include divide and conquer algorithms, graph algorithms, greedy algorithms, dynamic programming, NP-completeness and reductions.  The course goal is to provide a solid foundation in algorithms for students in preparation for a job in industry and more advanced courses.

Introduction to dynamics and control for robotic systems tailored for computer scientists. Concepts including ordinary differential equations, kinematics, and dynamics for common air and ground robotic systems will be introduced. Systems concepts such as step, impulse responses, Laplace Transform will be introduced. Feedback control via classical methods (e.g., Nyquist, Bode), PID, and modern state-space and observer-based design will be explored. Emphasis on implementation, and simulation on an aerial multicopter robot will help students visualize and evaluate learning and control design performance.