101 - 110 of 203 results for: EE

This course will introduce you to mixed signal design and the electronic design automation (EDA) tools used for it. Working in teams, you will create a chip with a digital deep neural network (DNN) accelerator and a small analog block using a modern design flow and EDA tools. The project involves writing a synthesizable C++ and a Verilog model of your chip, creating a testing/debug strategy for your chip, wrapping custom layout to fit into a standard cell system, using synthesis and place and route tools to create the layout of your chip, and understanding all the weird stuff you need to do to tape-out a chip. Useful for anyone who will build a chip in their Ph.D.

Electrical issues in the design of high-performance digital systems, including signaling, timing, synchronization, noise, and power distribution. High-speed signaling methods; noise in digital systems, its effect on signaling, and methods for noise reduction; timing conventions; timing noise (skew and jitter), its effect on systems, and methods for mitigating timing noise; synchronization issues and synchronizer design; clock and power distribution problems and techniques; impact of electrical issues on system architecture and design. Prerequisites: EE101A and EE108A. Recommended: EE114/214A.

The course focuses on the theory and algorithms underlying modern data compression. The first part of the course introduces techniques for entropy coding and for lossless compression. The second part covers lossy compression including techniques for multimedia compression. The last part of the course will cover advanced theoretical topics and applications, such as neural network based compression, distributed compression, and computation over compressed data.

(Formerly EE 376A.) Information theory was invented as a mathematical theory for communication but has subsequently found a broad range of applications. We study how to measure, represent, and communicate information effectively: from the foundational concepts of entropy and mutual information to the fundamental role they play in data representation, communication, inference, practical compression and error correction. As time allows, we cover relations and applications to other areas such as probability, statistics, learning and genomics.

The subject of reinforcement learning addresses the design of agents that improve decisions over time while operating within complex and uncertain environments. This first course of the sequence restricts attention to the special case of bandit learning, which focuses on environments in which all consequences of an action are realized immediately. This course covers desired agent behaviors and principled scalable approaches to realizing such behavior. Topics include learning from trial and error, exploration, contextualization, generalization, and representation learning. Motivating examples will be drawn from recommendation systems, crowdsourcing, education, and generative artificial intelligence. Homework assignments primarily involve programming exercises carried out in Colab, using the python programming language and standard libraries for numerical computation and machine learning. Prerequisites: programming (e.g., CS106B), probability (e.g., MS&E 121, EE 178 or CS 109), machine learning (e.g., EE 104/ CME 107, MS&E 226 or CS 229).

Many engineering applications require efficient methods to process, analyze, and infer signals, data and models of interest that are best described probabilistically. Building on a first course in probability (such as EE 178 or equivalent), this course introduces more advanced topics in probability such as concentration inequalities, random vectors and random processes, and explores their applications in statistics, machine learning and signal processing. Specific applications include hypothesis testing and classification; dimensionality reduction and generalization in machine learning, minimum mean square error estimation and Kalman filtering.

Digital communication is a rather unique field in engineering in which theoretical ideas have had an extraordinary impact on the design of actual systems. The course provides a basic understanding of the analysis and design of digital communication systems, building on various ideas from probability theory, stochastic processes, linear algebra and Fourier analysis. Topics include: detection and probability of error for binary and M-ary signals (PAM, QAM, PSK), receiver design and sufficient statistics, controlling the spectrum and the Nyquist criterion, bandpass communication and up/down conversion, design trade-offs: rate, bandwidth, power and error probability, coding and decoding (block codes, convolutional coding and Viterbi decoding).

Course focuses on how to build modern computing systems, namely notebooks, smartphones, and data centers, covering primarily their hardware architecture and certain system software aspects. For each system class, we cover the system architecture, processor technology, advanced memory hierarchy and I/O organization, power and energy management, and reliability. We will also cover topics such as interactions with system software, virtualization, solid state storage, and security. The programming assignments allow students to explore performance/energy tradeoffs when using heterogeneous hardware resources on smartphone devices.

Formulation and solution of sequential decision problems under uncertainty as a foundation for artificial intelligence, operations research, and economics. Finite-horizon, discounted, and average reward objectives. Optimization via value iteration, policy iteration, linear programming, and reinforcement learning algorithms. Semi-Markov decision processes. Multi-armed bandits and the Gittin's index theorem. Use of partial state information. Homework assignments involve a combination of analytic and computational exercises.