Electrical Engineering

An overview is given of a general communication link consisting of the three parts: transmitter, communication channel, and receiver. Examples of digital communication methods are introduced for realistic bit rates and noise levels. Some of the following applications are considered in the course: Mobile digital telephony (3G, EDGE, GSM), WLAN, modem, ADSL, digital TV, Bluetooth, navigation (GPS), surveillance systems.

This course examines signals and systems for modelling and analyzes a variety of engineering systems. It covers continuous‐ and discrete‐time Fourier analysis, Laplace Transform, interactions between signals and linear time invariant (LTI) systems, sampling theorem, differential and difference equations as LTI systems, and application examples in communications, control, and multimedia. 

This is a special studies course involving an internship with a corporate, public, governmental, or private organization, arranged with the Study Center Director or Liaison Officer. Specific internships vary each term and are described on a special study project form for each student. A substantial paper or series of reports is required. Units vary depending on the contact hours and method of assessment. The internship may be taken during one or more terms but the units cannot exceed a total of 12.0 for the year.

This course examines major topics in pattern recognition, particularly aspects of classification and decision. Students will gain effective pattern recognition tools with which to analyze the often vast amounts of diverse data in research applications. 

Topics include introduction to pattern recognition - machine perception - PR systems and design cycle, Bayesian decision theory for continuous features - Bayes Decision Rule - minimum-error-rate classification - classifiers, normal density, discriminant functions and discrete Bayesian decision theory - discriminant functions for the normal density - error probabilities and integrals - Bayes Decision Theory for discrete features, maximum-likelihood and Bayesian parameter estimation - Bayesian parameter estimation: Gaussian case - Bayesian parameter estimation: general theory - HMM, nonparametric techniques - density estimation - Parzen windows - nearest neighbor estimation (NN, k-NN) - fuzzy classification, linear discriminant functions i - linear discriminant functions and decision surfaces - generalized linear discriminant functions - minimizing the perceptron criterion function, relaxation procedures, linear discriminant functions ii - minimum square-error procedures - relation to Fishers linear discriminant - the Widrow-Hoff and Ho-Kashyap procedures - multicategory generalizations - ridge regression and its dual form [2] - classification error based method [2], model assessment and performance evaluation - bias, variance and model complexity [2] - model assessment and selection 2] - confusion matrix, error rates, and ROC [2] - statistical inference [2] - statistical errors [2], dimension reduction and feature extraction - principal component analysis - Fisher linear discriminant - nonlinear projections, support vector machines - introduction - SVM for pattern recognition [2] - linear support vector machines [2] - nonlinear support vector machines [2], multilayer neural networks - introduction - feedforward operation and classification - backpropagation algorithm - some issues in training neural networks - key ideas in classification, introduction to deep learning networks - convolutional neural networks (CNN) - autoencoders - deep belief networks - deep reinforcement learning - generative adversarial networks (GAN), algorithm-independent machine learning - introduction - bias and variance - resampling for classifier design - estimating and comparing classifiers - combining classifiers, unsupervised learning and clustering - mixture densities and identifiability - maximum-likelihood estimates - application to normal mixtures. 

Prerequisites: Linear Algebra, Probability, MATLAB, Python, or C-Programming Skills 

This course explores the basics of opto-electronics and photonics, which has many applications areas in information and communication technologies. By the end of the semester, students should have basic knowledge of (1) what light is, (2) how the basic property of light can be modeled, and (3) how light can be used for various applications. Topics include basics of electromagnetism, maxwell's equations, plane-wave solutions, polarization, EM waves in conductor, total internal reflection, interference, light incident on conductors, light incident on dielectric interface, multiple dielectric interface, interferometers, diffraction, metallic waveguides, dielectric waveguides, 2-D dielectric waveguides, optical fiber, waveguide devices, photons, interaction between light and matter, optical amplifiers, semiconductors, semiconductor lasers, single mode lasers, and photodetectors.

Prerequisite: Basic knowledge in electromagnetism

This course explores generative artificial intelligence (GAI) and its applications. Students gain a comprehensive understanding of generative models, including deep learning architecture, and probabilistic models. The course covers theoretical foundations and practical implementations of generative AI algorithms. Students also engage in hands-on projects to apply generative AI methods. Topics include introduction to generative AI (overview of generative modeling, brief history of GAI, applications of GAI), probability theory and information theory, parameters estimation, latent variable models, variational inference (introduction), variational autoencoders (VAEs) - autoencoders - variational autoencoders (VAE) - conditional VAE - VQ-VAE v1, v2, generative adversarial networks (GANs) - introduction to GANs - GAN training, issues and solution - generative model evaluation, GAN variants: DCGAN, CGAN, WGAN, ProGAN and Style-GAN, GAN applications: image manipulation and editing, diffusion-based generative models - DDPM - DDIM, diffusion-based generative models - classifier guidance DMs - classifier-free guidance DMs - cascaded DMs - latent DMs, autoregressive generative models - MADE, PixelNN, language generative models - Transformer - GPT family, multi-modal generative models - DALL-E (DALL-E 2 and DALL-E 3) - stable diffusion, flow-based generative models - RealNVP, GLOW.

Prerequisite: Solid understanding of machine learning and deep learning principles - Proficiency in programming - Familiarity with deep learning frameworks (e.g., TensorFlow, PyTorch)

This class provides fundamental understanding of energy conversion by use of power electronic devices. Students are expected to perform analysis and synthesis of power electronic systems after this course. Expected outcome includes: 1. Demonstrate the ability to analyze switching power converters in steady state using circuit averaging and determine DC voltages and currents 2. Be able to sketch current and voltage waveforms in a converter in steady state 3. Demonstrate the ability to size passive filtering components in converters such as inductors and capacitors to obtain a desired ripple performance 4. Demonstrate the ability to derive small-signal linearized models for switching converters 5. Demonstrate an understanding of the effects of negative feedback on converter operation 6. Demonstrate the ability to simulate switching converter using both switching models and averaged models via PSCPICE. 

Prerequisite: EEE2010 (Basic Circuit Theory) 

This course will introduce fundamental concepts and techniques in the content of remote sensing and image processing for Earth observation from space. The course starts by introducing core concepts in remote sensing (describing the processes by which images are captured by sensors mounted on satellite and airborne platforms and key characteristics of the acquired images). Then, fundamental methodologies for processing, analyzing, and visualizing remotely sensed imagery are introduced. Topics include representation of high-dimensional remote sensing images, time domain representations, filtering and enhancement. Practical applications will be provided throughout the course. Participants of this course will gain theoretical and practical knowledge on fundamental concepts and techniques for processing and analysis of remote sensing images acquired by Earth observation satellite and airborne systems.

This course provides an overview of traditional energy sources, electrical energy generation, transmission, distribution and utilization systems. It introduces the concepts of renewable energy sources, distributed renewable energy generation and smart-grid structure. The key issues of energy requirement in portable electronic computing system and wireless energy transfer are covered. The course requires students to take prerequisites.

Motion planning is a fundamental building block for autonomous systems, with applications in robotics, industrial automation, and autonomous driving. After completion of the course, students will have a detailed understanding of: Formalization of geometric, kinodynamic, and optimal motion planning; Sampling-based approaches: Rapidly-exploring random trees (RRT), probabilistic roadmaps (PRM), and variants; Search-based approaches: State-lattice based A* and variants; Optimization-based approaches: Differential Flatness and Sequential convex programming (SCP); The theoretical properties relevant to these algorithms (completeness, optimality, and complexity). Students will be able to: Decide (theoretically and empirically) which algorithm(s) to use for a given problem; Implement (basic versions) of the algorithms themselves; • Use current academic and industrial tools such as the Open Motion Planning Library (OMPL).

It provides a unified perspective on motion planning and includes topics from different research and industry communities. The goal is not only to learn the foundations and theory of currently used approaches, but also to be able to pick and compare the different methods for specific motion planning needs. An important emphasis is the consideration of both geometric and kinodynamic motion planning for the major algorithm types.