Machine Learning News
Kumar Avinava Dubey, a Ph.D. student in the Machine Learning Department, has been awarded an IBM Ph.D. Fellowship Award for the 2014-15 academic year.
The IBM Ph.D. Fellowship Awards Program is an intensely competitive worldwide program, which honors exceptional Ph.D. students who have an interest in solving problems that are important to IBM and fundamental to innovation in many academic disciplines and areas of study. Dubey’s research interests include statistical machine learning, non-parametric Bayesian methods, information retrieval and clustering. Fellows receive a stipend for one academic year. They also are matched with an IBM mentor according to their technical interests, and are strongly encouraged to participate in at least one internship at IBM while completing their studies.
Dubey earned a master’s degree in computer science at IIT Bombay and was a researcher for IBM Research India from 2009-2011, before coming to CMU as a Ph.D. student. His faculty adviser is Eric Xing, associate professor of machine learning, language technologies and computer science.
Christopher Warren, Mike Finegold and Cosma Shalizi have teamed up with Georgetown University’s Daniel Shore for the project “Six Degrees of Francis Bacon.” In the project — which has been supported by a Google Faculty Research Award — Warren, Finegold, Shalizi and Shore will combine technology and text to digitally reconstruct the early modern social network so that scholars and students all over the world will be able to collaborate, revise, curate and critique. Warren is an assistant professor of English, Finegold is a visiting assistant professor of statistics and Shalizi is an associate professor of statistics. Read more at http://sixdegreesoffrancisbacon.com/.
Assistant Professor Barnabás Poczós has been named the winner of a Yahoo! Academic Career Enhancement (ACE) award. This award is presented to top young faculty members in Yahoo!-relevant academic areas, and provides an unrestricted gift of $10,000 to support the faculty's research.
Alum, Rayid Ghani mentioned in NY Times article as Chief Scientist for the Obama campaign. Project Dreamcatcher - How cutting edge text analytics helped the Obama campaign determine voters' hopes and fears.
Read NY Times article
Best Paper Award at CIKM'12
Gelling, and Melting, Large Graphs by Edge Manipulation
Hanghang Tong, B. Aditya Prakash, Tina Eliassi-Rad, Michalis Faloutsos and Christos Faloutsos.
William F. Eddy, the John C. Warner Professor of Statistics, has chosen to accept Emeritus status. Since joining Carnegie Mellon in 1976, Eddy has worked in a variety of disciplines, with research covering theoretical probability, statistics and applied problems. His research has focused on the data generated by functional Magnetic Resonance Imaging (fMRI), a technique used by cognitive neuroscientists to chart brain activity. A distinguished scholar, Eddy has published more than 100 research papers and authored or edited 20 books and monographs. Eddy will be honored at a reception at 4:30 p.m., Friday, Nov. 2 in the Newell Simon Hall Atrium.
On Oct. 23, Joseph B. (Jay) Kadane, the Leonard J. Savage University Professor of Statistics and Social Sciences, emeritus, was named Statistician of the Year by the Chicago Chapter of the American Statistical Association for his outstanding contributions to the field of statistics.
NSF Big Data Initiative Awards
The National Science Foundation, with support from the National Institutes of Health, awarded two grants totaling more than $1.7 million to Carnegie Mellon researchers to support new big data research projects on Oct. 3.
Christos Faloutsos, a professor in the Machine Learning Department, received an $894,892 award to develop theory and algorithms to tackle the complexity of language processing, and to develop methods that approximate how the human brain works in processing language. Faloutsos' co-principal investigator is Professor Tom Mitchell of Machine Learning.
Aarti Singh, assistant professor in Machine Learning, was awarded $820,000 to develop new statistical and algorithmic machine learning approaches that would benefit the fields of physics, psychology, economics, epidemiology, medicine, and social network analysis. Singh's co-PIs are Timothy Verstynen, assistant professor of psychology in the Center for the Neural Basis of Cognition, and Barnabás Póczos, assistant professor in machine learning. Read the NSF announcement.
Paper highlighted in Journal of the American Statistical Association (JASA)
Qirong Ho, Ankur Parikh and Eric Xing of the Machine Learning Department propose “A Multiscale Community Blockmodel for Network Exploration” that allows investigators to infer these phenomena from a set of observed network interactions. Ho et al. develop a stochastic model for partitioning the units in the network, say species, in a hierarchically organized tree. Each species’ interactions are governed by a multiscale membership vector that describes the species likelihood of interacting with species at different levels of the hierarchical tree.
Finally, a probability model that links the hierarchical tree and the membership vectors to observed network connections can be used to infer the parameters of the model. The authors demonstrate the approach on a network describing the predator-prey relationships among a collection of 75 species of grass-feeding wasps and their parasites.
Manuela Veloso, the Herbert A. Simon professor, was recently profiled in Wired Magazine as the "World's Most Wired Robot Master" for her CoBots (short for "collaborative robot"). Veloso aims to make robots that can anticipate and handle physical unknowns. Obstacle avoidance may not seem very glamorous at first glance — most people want their robots to shoot lasers or cry "danger!" — but it's a linchpin of real-world robotics and one of Veloso's key areas of focus. Read the article.
What Makes Paris Look Like Paris? CMU Software Uncovers Stylistic Core
Paris is one of those cities that has a look all its own, something that goes beyond landmarks such as the Eiffel Tower or Notre Dame. Researchers at Carnegie Mellon University and INRIA/Ecole Normale Supérieure in Paris have developed visual data mining software that can automatically detect these sometimes subtle features, such as street signs, streetlamps and balcony railings, that give Paris and other cities a distinctive look.
Our data mining technique was able to go through millions of image patches automatically — something that no human would be patient enough to do," said Alexei Efros, who collaborated with colleagues including Abhinav Gupta, assistant research professor of robotics, and Carl Doersch, a Ph.D. student in the Machine Learning Department. Read More
Associate Professor, Eric Xing to direct dual-degree Computer Science Masters program with Tsinghua Universities. Read more
Christos Falousos receives honorary degree from The Aristotle University of Thessaloniki, the largest university in Greece. The title of Doctor Honoris Causa was conferred to Faloutsos during a May 30 convocation, one of a number of events this month that mark the 20th anniversary of Aristotle University's Department of Informatics. Read the full story.
Emma Brunskill is announced a Wimmer Faculty Fellow. The Eberly Center for Teaching recently announced its 2012-2013 Wimmer Faculty Fellows. Made possible by a grant from the Wimmer Family Foundation, the awards are designed for junior faculty members interested in enhancing their teaching through concentrated work designing or re-designing a course, innovating new materials or exploring a new pedagogical approach. Fellows work in close collaboration with Eberly Center colleagues and receive a stipend to acknowledge the work it takes to improve one's effectiveness as an educator.
Emma Brunskill, assistant professor of computer science at Carnegie Mellon University, is one of seven recipients this year of a prestigious Microsoft Research Faculty Fellowship, which recognizes pioneering young academic researchers working in key areas of computer science. Read more
Jaime Carbonell, Allen Newell Professor of Computer Science and director of the Language Technologies Institute in the School of Computer Science, has received the elite distinction of University Professor, the highest academic accolade a faculty member can achieve at Carnegie Mellon. Read more
John Lafferty & alum, Risi Imre Kondor to receive the ICML Test of Time Award for their paper "Diffusion Kernels on Graphs and Other Discrete Input Spaces". ICML 2002: 315-322. To be presented at the ICML 2012 Conference in Edinburgh, Scotland.
Carlos Guestrin, Associate professor in the machine learning & computer science departments, along with Seth Goldstein in the computer science department have created Flashgroup, a startup featuring new technology. Flashgroup allows people who couldn't ordinarily find each other to connect in real time — then disband just as easily when the 'flash' is over. Read more
Eric Xing, Associate professor of machine learning, has been awarded an IBM Open Collaboration Research (OCR) Award to develop novel and scalable ways to use very large data sets to search for associations between genetic variants and traits such as major diseases. Xing will receive $300,000 over two years for his project, “Scalable Genome-wide Association Studies: Towards a pipeline for large-scale personal genomics.” The OCR is the highest research award given by IBM to university faculty. Read more.
Pegasus graph mining system developed by SCS PhD Students, U. Kang and Polo Chau, is chosen by Microsoft to be included in their 'hadoop' distribution. Pegasus has already received two Best Paper Awards, and an Open Source Software Award and is capable of analyzing huge, disk-resident graphs and social networks.Both students are advised by Christos Faloutsos.
New research from Carnegie Mellon University's Marcel Just provides an explanation for some of autism's mysteries, giving scientists clear targets for developing intervention and treatment therapies. Read more.
Stephen Fienberg, the Maurice Falk University Professor of Statistics and Social Science, has received a research grant from the National Science Foundation (NSF)-Census Bureau Research Network for a project on "Data Integration, Online Data Collection and Privacy Protection for the 2020 Census." Fienberg and his team will conduct research on three basic issues of interest related to collecting census data: privacy, costs and response rates. They will address the practical problems of ensuring confidentiality and privacy while still producing useful statistics for public and private purposes. Read more.
Carnegie Mellon Performs First Large-Scale Analysis Of “Soft” Censorship of Social Media in China. The study by Noah Smith, associate professor in the Language Technologies Institute (LTI); David Bamman, a Ph.D. student in LTI; and Brendan O’Connor, a Ph.D. student in the Machine Learning Department, appears in the March issue of First Monday, a peer-reviewed, online journal. Read more.
Ziv Bar-Joseph, associate professor of computational biology and machine learning, and graduate student Guy Zinman are among the authors of a paper in the journal Nature that for the first time links overexpression of a gene called sirtuin 6 to increased life span in mammals, specifically mice. Researchers who study aging have been intrigued by the large family of sirtuin genes and their proteins ever since they were linked to longevity in yeast.
Prof. Christos Faloutsos will be awarded an honorary Ph.D. degree by the Aristotle University of Thessaloniki, Greece. The ceremony will take place in June 2012, in Thessaloniki.
FAWM (February Album Writing Month) Founder, Burr Settles mixes song writing and machine learning. The premise of the project is this: You sign up and create a profile at the FAWM website, www.fawm.org. Then, during the month of February, your goal is to write 14 songs. You record and share them on the website, and give and receive feedback with other participants ("fawmers"). Read more
Tom Mitchell, Department Head of Machine Learning and Manuela Veloso will be among the presenters at the World Economic Forum, Jan. 25-29 in Davos, Switzerland. The WEF is a prestigious independent international organization committed to improving the state of the world by engaging business, political and academic leaders to shape global, regional and industrial agendas. Read more
Noah Smith, associate professor in the Language Technologies Institute, has been awarded a three-year Finmeccanica Career Development Chair. His research interests include statistical natural language processing, especially unsupervised methods, machine learning for structured data and applications of natural language processing. An endowment by the Italian conglomerate Finmeccanica funds two chairs that support outstanding young faculty members in the School of Computer Science. The other chair is currently held by Alexei Efros, associate professor of Computer Science and Robotics.
Robert F. Murphy, director of the Ray and Stephanie Lane Center for Computational Biology, will lead a multidisciplinary team of researchers that will collaborate with scientists at the Baylor College of Medicine and Yale University. The ultimate dream, Murphy said, is to develop what Ion Torrent Founder and CEO Jonathan M. Rothberg dubbed "doctor in a box," software. Doctor-in-a-box would take a patient's DNA sequence and use it to diagnose disease, identify a patient's susceptibility to disease, and predict which therapies might be most effective or cause the fewest side effects. Read more
The Chinese Academy of Sciences has named Manuela Veloso, the Herbert A. Simon Professor of Computer Science at Carnegie Mellon University, as an Einstein Chair Professor for 2012. She is one of 20 prominent international scientists so honored. As an Einstein Chair Professor, Veloso will present a lecture at the University of Science and Technology of China, a national research university in Hefei, China, and at another Chinese university. For more information, see CMU’s announcement.
Ramayya Krishnan, Heinz College Dean, related faculty in the Machine Learning Dept., has received the Institute for Operations Research and the Management Sciences (INFORMS) Information Systems Society Distinguished Fellow Award in recognition of his intellectual contributions to the information systems discipline. His current research projects investigate risk management in business process design and in information security, social network analysis in settings ranging from call data records to knowledge sharing communities, consumer behavior in e-business settings, and the design of policies that take into account the competing needs of promoting data access and protecting privacy.
Ann Lee, Associate Professor in Statistics & Co-Director of the Machine Learning graduate programs, is awarded the Estella Loomis McCandless Career Development Chair.
ISMB Best Paper Award in Translational Bioinformatics
Ankur Parikh, W. Wu, R. Curtis and E. P. Xing for " TREEGL: Reverse Engineering Tree-evolving Gene Networks Underlying Developing Biological Lineages".
Leila Wehbe and Alona Fyshe, PhD candidates in the Machine Learning Department are among the inaugural recipients of Rothberg Research Awards in Human Brain Imaging. They will use their share of the $100,000 prize to use machine learning, fMRI and magnetoencephalography (MEG) to model the neural processing of a word to understand how it is affected by semantic context. For more information, see CMU’s announcement.
Alumnus Daniel Wilson's first novel, "ROBOPOCALYPSE," has been released. Wilson's seventh book tells the story of a group of people trying to survive after technology has turned against them.
Read more: http://www.danielhwilson.com/
ICML 2011 Best Paper Award
Kevin Waugh, Brian Ziebart and Drew Bagnell for "Computational Rationalization: The Inverse Equilibrium Problem".
ICML 2011 Test of Time Award for the 10 Year Best Paper
John Lafferty , Andrew McCallum, and Fernando C. N. Pereira, "Conditional Random Fields: Probabilistic Models for Segmenting and Labeling Sequence Data".
SCS team won the best application paper award at PAKDD 2011
"Spectral Analysis for Billion Scale Graphs: Discoveries and Implementation" by U Kang, Brendan Meeder and Christos Faloutsos, PAKDD'11, Shenzhen China. The paper shows how to do accurate spectral analysis on large graphs using 'hadoop', and showed how to use the results for anomaly detection. PAKDD is one of the top data mining conferences.
Professor Avrim Blum will be awarded the Herbert A. Simon Award for Teaching Excellence in Computer Science at the SCS Diploma Ceremony. http://www.cmu.edu/homepage/computing/2011/spring/avrim-blum.shtml
Jiashun Jin, Associate Professor of Statistics, has received a fellowship with the Institute of Mathematical Statistics (IMS). IMS selects fellows to honor outstanding research and personal contributions that keep IMS in a leading role in the field of statistics and probability. Jin's current research is in large-scale inference and massive-data analysis, which are frequently found in many scientific areas, such as genomics, astronomy, functional magnetic resonance imaging and image processing.
Professor William Cohen has been elected the next President of the International Machine Learning Society. The IMLS is the professional society that organizes the annual International Conference on Machine Learning (ICML).
Daniel Neill, Assistant Professor, Heinz School and Machine Learning Dept. and Alumni, Jure Leskovec named "10 to Watch" by IEEE Intelligent Systems magazine, which recognizes 10 outstanding young researchers in AI.
NSF Fellowship awarded to Ankur Parikh, ML PhD student working with Eric Xing.
IBM Fellowship awarded to Xi Chen, ML PhD student working with Jaime Carbonell.
Open House for Accepted PhD Students
Manuela Veloso and Larry Wasserman have been elected 2010 American Association for the Advancement of Science (AAAS) Fellows for their distinguished efforts to advance science or its applications. They will be honored at the AAAS Fellows Forum, Feb. 19 at the 2011 AAAS Annual Meeting in Washington, D.C. Read more
Noah Smith Receives NSF CAREER Award
The National Science Foundation has awarded Noah A. Smith, assistant professor in the Language Technologies Institute and Machine Learning Department, a five-year, $550,000 Faculty Early Career Development (CAREER) Award to study flexible statistical learning algorithms for natural language processing. Smith's research focuses on computational models of human language: formal aspects, learning such models from data, and applying them to problems like translation and social media analysis.
Fruit Fly Nervous System Provides New Solution To Fundamental Computer Network Problem
Ziv Bar-Joseph is working on a biologically derived method that could help organize wireless sensor networks. The fruit fly has evolved a method for arranging the tiny, hair-like structures it uses to feel and hear the world that's so efficient a team of scientists in Israel and at Carnegie Mellon University says it could be used to more effectively deploy wireless sensor networks and other distributed computing applications. News Release
Burr Settles, a post-doctoral fellow in the Machine Learning Department, has developed Titular and LyriCloud, two online creativity tools for songwriters. Titular suggests song titles, while LyriCloud makes lyrical suggestions based on words selected by the user. The motivation for Settles to create Titular and LyriCloud last year was February Album Writing Month (FAWM, http://fawm.org), an international songwriting event that he helped launch in 2004. FAWM challenges participants ("fawmers") to compose 14 new works of music — roughly an album's worth — in only 28 days. Last year, 4,000 people registered for FAWM and generated more than 10,000 new songs. Titular and LyriCloud, along with tools related to song structure and plot lines, were made available on the FAWM website to help participants in achieving their 14-song goals. Read more: http://www.cmu.edu/news/archive/2011/January/jan20_songtool.shtml
Michael J. Tarr, co-director of the Center for the Neural Basis of Cognition (CNBC), has received a professorship for his contributions to further understanding the brain and mind. Tarr was named the George A. and Helen Dunham Cowan Professor of Cognitve Neuroscience. Tarr studies the neural, cognitive and computational mechanisms underlying visual perception and cognition. He is particularly interested in how we effortlessly learn, remember, and visually identify both human faces and objects, as well as how these mechanisms of vision interact with our other senses, thoughts and emotions. See CMU's announcement.
ML and CS Professor Christos Faloutsos and his brothers, Michalis (University of California-Riverside) and Petros (UCLA), were honored with the Test of Time Award for their 1999 paper, "On the Power Law Relationships of the Internet Topology" by SIGCOMM, the premier computer communications conference. News Release
Save The Census: ML and Statistics Professor, Stephen E. Fienberg warns that demographic data provided is too valuable to lose. For the full news release visit http://www.cmu.edu/news/archive/2010/August/aug26_savethecensus.shtml
Best Research Paper: Innovative Contribution at KDD-2010 goes to, "Connecting the Dots between News Articles" by Dafna Shahf and Carlos Guestrin. Congratulations!
Recognized Finalist goes to "Discriminative Topic Modeling based on Manifold Learning" by Stephen E. Fienberg and Seungil Huh. Congratulations!
Computer Science and ML Professor, Christos Faloutsos will receive the 2010 Innovation Award from the Association for Computing Machinery's Special Interest Group on Knowledge Discovery and Data Mining (SIGKDD) at the KDD 2010 international conference July 25-28 in Washington, DC. Faloutsos' cross-disciplinary works on power-law graphs, fractal-based analysis, time series, multimedia and spatial indexing is among the most referenced in industry and academic publications. He and his students also have devised software for identifying accomplices on Internet auction sites that help others perpetrate fraud.
Best Paper Award at PAKDD 2010 goes to "OddBall: Spotting Anomalies in Weighted Graphs"
by Leman Akoglu, Mary McGlohon and Christos Faloutsos and gives fast algorithms to spot strange nodes in large social networks. PAKDD is one of the top data mining conferences. The paper was selected among 412 submissions, and 42 accepted full-papers.
Machine Learning Best Paper Awards
Best Student Paper (COLT 2010)
"Theoretical Justification of Popular Link Prediction Heuristics"
Purnamrita Sarkar, Deepayan Chakrabarti, Andrew Moore
Best Paper (ICML 2010)
"Hilbert Space Embeddings of Hidden Markov Models"
Le Song, Byron Boots, Sajid Siddiqi, Geoff Gordon, and Alex Smola
Runnerup Best Student Paper (ICML 2010)
"Modeling Interaction via the Principle of Maximum Causal Entropy"
Brian Ziebart, Anind Dey, Drew Bagnell
Manuela Veloso, the Herbert A. Simon Professor in the Computer Science Department, has been elected president of the Association for the Advancement of Artificial Intelligence (AAAI). Veloso will be the fourth AAAI president from Carnegie Mellon, following Allen Newell, Raj Reddy and Tom Mitchell.
William F. Eddy, the John C. Warner Professor of Statistics, will complete his second three year term as chairman of the Committee on National Statistics (CNSTAT) at the National Academies on June 30. CNSTAT is an independent and objective resource for evaluating and improving the work of the highly decentralized U.S. federal statistical system. Eddy is the only person to have chaired both statistics committees at the National Academies, having previously been chairman of the Committee on Applied and Theoretical Statistics.
Philosophy Professor Richard Scheines has been reappointed head of the Department of Philosophy, a position he has held since 2005. Scheines has been faculty member since 1990 and has additional appointments in ML and HCII. His research focuses on causal discovery, the philosophy of social science and educational technology and online courses.
Noah Smith, an assistant professor of Language Technologies, and his team of researchers found that sentiments expressed in a billion Twitter messages were similar to those of well-established public opinion polls. "With seven million or more messages being tweeted each day, this data stream potentially allows us to take the temperature of the population very quickly," Smith said. "The results are noisy, as are the results of polls. Opinion pollsters have learned to compensate for these distortions, while we're still trying to identify and understand the noise in our data. Given that, I'm excited that we get any signal at all from social media that correlates with the polls." For more on the research, which will be presented May 25 at the Association for the Advancement of Artificial Intelligence's International Conference on Weblogs and Social Media in Washington, DC, visit http://www.cmu.edu/news/archive/2010/May/may11_twitterpolls.shtml