Machine Learning

SUMMARY

Featured Research: Firebird - Predicting Fire Fisk in Atlanta

Existing fire inspection practices rely mostly on intuition. Georgia Tech reseachers (Shang-Tse Chen, Wenwen Zhang, Dr. P. Chau, Dr. B. Dilkina) and others created a data-driven framework for helping fire departments identify new sites requiring inspection and prioritize fire inspections. This research was selected as the "Best Student Paper Award" runner-up at 2016 ACM SIGKDD conference.
View Project Page
Featured Research: Modeling Intertwined Processes in Online Social Networks

In online social networks (e.g. Facebook, Twitter, etc), the processes of Information diffusion and network evolution are highly intertwined. These two have been dominantly studied separately, however. Georgia Tech Researchers (Mehradad F., Yichen W., Shuang L., Dr. Le Song) and others developed a novel model, referred to as COEVOLVE, in order to model the co-evolutionary dynamics of these two processes.
READ MORE
Featured Research: Automating the Construction of Agent-based Models of Fish Schooling Behavior

Agent-based models (ABM) are effective for validating theoretical models used for interpreting animal behaviors. Georgia Tech researchers (Brian Hrolenok and Dr. Tucker Balch) developed an automated tool for constructing such models in order to help biologists focus more time on analysis. Picture shown is a model of fish schooling behavior, used for empirical evaluation in the paper.
Read Paper
Featured Research: Object Recognition - Improving Indoor Semantic Labeling

Traditionally, indoor semantic labeling is done using pixels or superpixels. Georgia Tech Researchers (Steven Hickson, Dr. I. Essa, Dr. H. Christensen) developed a novel mechanism for achieving better segmentation using hierarchy of superpixels.
Read Paper
Featured Research: Exploring Gesture Controls for Visualizing Time-Evolving Graphs

Undergraduate researchers, Samuel Clarke and Nathan Dass, worked with Dr. Polo Chau to develop gestural interaction techniques for exploring time-evolving graph data. The techniques proved successful for exploring graphs with approximately 2500 time-based edges in real time.
View Video
Featured Research: Improving Gaussian Process Emulators

Gaussian process models are useful emulators for computer experiments but are computationally expensive. Georgia Tech Researchers (Chih-Li Sung and Dr. Benjamin Haaland) studied subsample locations in local Gaussian process approximation that can potentially save substantial computations while retaining emulation accuracy.
Read Paper
Featured Research: Discovering Picturesque Highlights from Egocentric Vacation Video

Georgia Tech Researchers (Dr. D. Castro, Dr. V. Bettadapura, Dr. I Essa) developed novel techniques for automatically determining aesthetic features and ranking video frames based on photographic qualities to generate highlights.
Read Paper
Featured Research: The Sketchy Database

Georgia Tech Researchers (Patsorn Sangklov, Nathan Burnell, Cusuh Ham and Dr. James Hayes) developed the Sketchy Database for training convolutional networks that embed sketches and photographs in a common feature space. The database can be a benchmark for fine-grained image retrieval as well as sketch and image synthesis.
Project Page
Featured Research: Clustering Continuous Live-Stream Documents

Georgia Tech Researchers (Nan Du, Mehrdad Farajtabar and Dr. Le Song) and others designed a novel random process, referred as the Dirichlet Hawkes process, for taking in both textual patterns and temporal dynamics of arriving patterns in order improve live-stream document clustering.
Read More
Featured Research: Robot Programming Paradigm for Modeling Ant Behavior

In biology, concise and testable models of animal behavior is desired for supporting the cycle of hypothesis generation and animal model testing. Georgia Tech Researchers (Yu-Ting Yang, Dr. A Quitmeyer, Brian Hrolenok, Harry Shang, Dihn Bao Nguyen and Dr. Tucker Balch) developed a robot programming paradigm, including programming libraries, for effective modeling of ant behaviors.
Read Paper

ABOUT ML@GT

Machine learning aims to produce machines that can learn from their experiences and make predictions based on those experiences and other data they have analyzed. The Center for Machine Learning at Georgia Tech (ML@GT) is an Interdisciplinary Research Center that is both a home for thought leaders and a training ground for the next generation of pioneers.

The field of machine learning crosses a wide variety of disciplines that use data to find patterns in the ways both living systems, such as the human body and artificial systems, such as robots, are constructed and perform. Whether it’s being applied to analyze and learn from medical data, or to model financial markets, or to create autonomous vehicles, machine learning builds and learns from both algorithm and theory to understand the world around us and create the tools we need and want.

Research

People

Seminar

Featured Blogs

Welcome to ML@GT by the Director

Welcome to ML@GT, a new interdisciplinary research center at Georgia Tech dedicated to the study and application of machine learning—one of the fastest developing research areas impacting computing, engineering, sciences and many other disciplines.

Welcome new faculty member Devi Parikh to School of Interactive Computing

Please join us in welcoming a new faculty member Devi Parikh in IC. Devi Parikh is an assistant professor at the School of Interactive Computing at Georgia Institute of Technology starting Fall 2016. From 2013 to 2016 she was an Assistant Professor in the Bradley Department of Electrical and Computer Engineering at Virginia Tech.

Welcome to new faculty Mayya Zhilova in Math

We would like to welcome Mayya Zhilova, who joined the School of Mathematics at the Georgia Institute of Technology in Fall 2016. Mayya's primary research interests lie in the areas of mathematical statistics and probability theory, particularly in statistical inference for complex high-dimensional data.

Georgia Institute of Technology

ML@GT

Search form