341 - 350 of 366 results for: CS

Most successful machine learning algorithms of today use either carefully curated, human-labeled datasets, or large amounts of experience aimed at achieving well-defined goals within specific environments. In contrast, people learn through their agency: they interact with their environments, exploring and building complex mental models of their world so as to be able to flexibly adapt to a wide variety of tasks. One crucial next direction in artificial intelligence is to create artificial agents that learn in this flexible and robust way. Students will read and take turns presenting current works, and they will produce a proposal of a feasible next research direction. Prerequisites: CS229, CS231N, CS234 (or equivalent).

How can we understand intelligent behavior as computation? This course introduces probabilistic programming as a tool for cognitive modeling. We will use probabilistic generative models to explain aspects of human and artificial cognition. Topics will be drawn from causal and social reasoning, concept learning, and hierarchical abstraction.

How can we understand natural language use in computational terms? This course surveys probabilistic models for natural language semantics and pragmatics. It begins with an introduction to the Rational Speech Acts framework for modeling pragmatics as social reasoning. It then explores a variety of phenomena in language meaning and usage. Probabilistic programming will be used as a precise and practical way to express models.

Interdisciplinary seminar focusing on understanding how computations in the brain enable rapid and efficient object perception. Covers topics from multiple perspectives drawing on recent research in Psychology, Neuroscience, and Computer Science. Emphasis on discussing recent empirical findings, methods and theoretical debates in the field.

Computer vision -- the study of how to design artificial systems that can perform high-level tasks related to image or video data (e.g. recognizing and locating objects in images and behaviors in videos) -- has seen recent dramatic success. In this course, we seek to give education and social science researchers the know-how needed to apply cutting edge computer vision algorithms in their work as well as an opportunity to workshop applications. Prerequisite: python familiarity and some experience with data.

Techniques and algorithms for creating effective visualizations based on principles from graphic design, visual art, perceptual psychology, and cognitive science. Topics: graphical perception, data and image models, visual encoding, graph and tree layout, color, animation, interaction techniques, automated design. Lectures, reading, and project. There are no official prerequisites for the class, but familiarity with the material in CS147, CS148 and CS142 is especially useful. Most important is a basic working knowledge of, or willingness to learn, web- programming, especially JavaScript, Vega-Lite and D3.js.

Digital photography and basic image processing, convolutional neural networks for image processing, denoising, deconvolution, single pixel imaging, inverse problems in imaging, proximal gradient methods, introduction to wave optics, time-of-flight imaging, end-to-end optimization of optics and imaging processing. Emphasis is on applied image processing and solving inverse problems using classic algorithms, formal optimization, and modern artificial intelligence techniques. Students learn to apply material by implementing and investigating image processing algorithms in Python. Term project.

Intermediate level, emphasizing physically based simulation techniques for computer animation and synchronized sound synthesis. Topics vary from year to year, but include the simulation of acoustic waves, and integrated approaches to visual and auditory simulation of rigid bodies, deformable solids, collision detection and contact resolution, fracture, fluids and gases, and virtual characters. Students will read and discuss papers, and do programming projects. Prerequisite: None. Recommended: Prior exposure to computer graphics and/or scientific computing.

Contents of this course vary with each offering. Past offerings have included geometric matching, surface reconstruction, collision detection, computational topology, differential geometry for computer scientists, computational symmetry and regularity, data-driven shape analysis, and non-Euclidean methods in machine learning.