News featuring Bert Huang

Congratulations to our Ph.D. and master’s degree Spring graduates at the Discovery Analytics Center!

DAC graduates include (left to right): Xuchao Zhang with advisor Chang-Tien Lu in the National Capital Region; and Elaheh Raisi with Bert Huang and Yufeng Ma hooded by Ed Fox, both in Blacksburg.

The Discovery Analytics Center is pleased to announce that five of their Ph.D. and four of their master’s degree students celebrated graduation from Virginia Tech last weekend at Commencement ceremonies in Blacksburg and in the National Capital Region.

“It is always bittersweet to bid our students farewell, but we wish them all the best. We know and appreciate how hard they have worked to achieve the high goals they set for themselves and look forward to following their successful careers in academia and industry,” said Naren Ramakrishnan, the Thomas L. Phillips Professor of Engineering and director of the center.

 

Ph.D. graduates

Sorour E. Amiri, advised by B. Aditya Prakash, received a Ph.D. in computer science.
Her research interests are large-scale graph mining, data mining, and applied machine learning and the title of her dissertation is “Task-specific Summarization of Networks: Optimization and Learning.” She is joining the Google search ad team.

Minghan Chen, co-advised by Layne Watson, received a Ph.D. in computer science. Her research interest is computational cell biology and her dissertation title is “Stochastic Modeling and Simulation of Multiscale Biochemical Networks.” She joins the Computer Science Department at Wake Forest University as assistant professor.

Yufeng Ma, co-advised by Weiguo (Patrick) Fan and Edward Fox, received a Ph.D. in computer science. Ma’s research interests are computer vision, Natural Language Processing (NLP), and deep learning and his dissertation title is “Going Deeper with Images and Natural Language.” Ma is joining Verizon Media (Yahoo! Research) as a research scientist focusing on personalized recommendations.

Elaheh Raisi, advised by Bert Huang, received a Ph.D. in computer science. Her research interests are machine learning, weakly supervised learning, and computational social science and her dissertation title is “Weakly Supervised Machine Learning for Cyberbullying Detection.”

Xuchao Zhang, advised by Chang-Tien Lu, received a Ph.D. in computer science. His research interests are data mining, machine learning, and Natural Language Processing (NLP) and his dissertation title is “Scalable Robust Models Under Adversarial Data Corruption.”  Zhang joins NEC Labs America as a researcher. In that position he will work to fully understand the dynamics of big data from complex systems; retrieve patterns to profile them; and build innovative solutions to help end user managing those systems.

Master’s graduates

Raja Venkata Satya Phanindra Chava, advised by Edward Fox, received a master of engineering degree. His research interests are text summarization using deep learning and Natural Language Processing (NLP) and his project title is “Natural Language Processing Techniques for Comprehending Legal Depositions.” Chava is joining Walmart in Reston, Virginia, as a software engineer and will work on big data analysis to manage the supply chain and personalize the customer’s shopping experience.

Supritha B. Patil, advised by Edward Fox, received a master of science degree in computer science. Patil’s research interest is Natural Language Processing (NPL) and her thesis title is “Analysis of Moving Events Using Tweets.” She will be working as a software developer.

Adithya Upadhya, advised by Edward Fox, received a master’s in computer science. His research interests are machine learning and high performance computing and his project title is “A General Web Platform Summarizing Text and Documents.”

Xinfeng Xu, advised by B. Aditya Prakash, received a master’s degree in computer science. His research focused on modeling and predicting incidence and the title of his thesis is “Modeling and Predicting Incidence: Critical Systems Failures and Flu Infection Cases.” He also received the 2019 MS Research Award from the Department of Science. Xu is also a Ph.D. student in physics in the College of Science and will continue his research in that field.

 

 

 

 


Amazon Research Award supports developing algorithms that tackle unfairness in recommendation engines

Bert Huang, DAC faculty member and assistant professor of computer science

Why would a recommendation engine not suggest computer science classes to a female college student interested in that field of study?

According to Bert Huang, assistant professor of computer science in the College of Engineering and a faculty member at the Discovery Analytics Center, there are a few reasons. The engine may have trained from data representing the existing gender imbalance in computer science, unfair patterns may have inadvertently emerged from the mathematical nature of its learning algorithm and model, or there may be a less-visible or harder-to-detect process in place.  Click here to read more about Bert’s work.


DAC faculty and students attend AAAI-19 conference to discuss their research

You Lu (left) and Chidubem Arachie (right), both DAC Ph.D. student in computer science, presenting their posters.

Discovery Analytics Center faculty Bert Huang and Chandan Reddy and two Ph.D. students were in Honolulu, Hawaii, last week, sharing their research with attendees at the Thirty-Third AAAI Conference on Artificial Intelligence.

Chidubem Arachie and You Lu, both in the Department of Computer Science, presented spotlight talks on studies they collaborated on with Huang, who is their advisor. The studies were also included in a poster session.

Arachie’s presentation was on Adversarial Label Learning (ALL), a method they introduced to train robust classifiers when access to labeled training data is limited. ALL trains a model without labeled data by making use of weak supervision to minimize the error rate for adversarial labels, which are subject to constraints defined by the weak supervision. Their study demonstrated that their method is robust against weak supervision signals that make dependent errors. Their experiments confirm that ALL is able to learn models that outperform the weak supervision and baseline models. ALL is also capable of directly training classifiers to mimic the weak supervision.

Lu presented Block Belief Propagation for Parameter Learning in Markov Random Fields. In this paper, the researchers developed block belief propagation learning (BBPL) for training MRF and theoretically proved that BBPL has a linear convergence rate and that it converges to the same optimum as convex BP. Their experiments show that, since BBPL has much lower iteration complexity, it converges faster than other methods that run truncated or complete inference on the full MRF each learning iteration.

Reddy gave an invited talk at the AAAI conference workshop on health intelligence, where he discussed ways to create natural language interfaces for both clinicians and patients and enable them to ask some basic questions on medical knowledge bases and clinical databases. This work, he said, will efficiently and effectively provide answers to questions and results to queries in the biomedical domain.


DAC Student Spotlight: Elaheh Raisi

DAC Ph.D. student, Elaheh Raisi

Elaheh Raisi’s enthusiasm for math dates back to high school. So it was not surprising when Raisi chose applied mathematics as her major at the Amirkabir University of Technology -Tehran Polytechnic.

“I realized early on that mathematics is essential for many practical sciences,” said Raisi. “My aim was to gain a strong knowledge of mathematics that I could use in problem solving.”

During her freshman year Raisi concentrated on mathematics and programming-related courses but after taking some computer science classes, she developed an interest in artificial intelligence. She earned a master’s degree in artificial intelligence at the Science and Research branch of the Islamic Azad University.

Raisi chose to pursue a Ph.D. in computer science at Virginia Tech because of “the abundance of advanced research resources and facilities in the field of data mining, a supporting environment at a prestigious university, and outstanding professors.” Now, a fifth-year Ph.D. candidate, Raisi is advised by Professor Bert Huang and works on cyberbullying detection on social media in his Machine Learning Laboratory.

To address the computational challenges associated with designing automated, data-driven machine learning approaches for harassment-based cyberbullying detection Raisi and Huang have developed a weakly supervised framework, which is specialized for cyberbullying detection,” she said.

The framework consists of two learning algorithms to improve predictive performance by taking into account not only language, but also social structures. One learner identifies bullying incidents by examining the language content in the message; another learner considers social structure to discover bullying. Intuitively, each learner is using different body of information. The learning algorithm tries to make them eventually agree whether social interactions are bullying.

“Our research is geared toward a very important topic in any online automated harassment detection: fairness against particular targeted groups including race, gender, religion, and sexual orientations,” said Raisi. “Our goal is to decrease the sensitivity of models to language describing particular social groups.”

For their research, Raisi and Huang won a best paper award at the IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM) 2017.

They also won a best paper award at the Learning with Limited Labeled data (LLD) workshop at NIPS, 2017 for including deep learning methodologies (word and node embedding) into their framework.


NIPS Conference 2017 showcases work from DAC Ph.D. students

 

 

 

 

 

 

 

 

 

 

 

 

A group of Ph.D. students from the Discovery Analytics Center headed with their faculty advisors to Long Beach, California, last week to present papers and posters at the 2017 Conference on Neural Information Processing Systems (NIPS). One of the workshop papers was distinguished with a Best Paper Award and two of the students received NIPS Travel Awards.

2017 marks the 31st year for the international multi-track machine learning and computational neuroscience conference includes invited talks, demonstrations, symposia, and oral and poster presentations of refereed papers, and workshops.

The Women in Machine Learning Workshop was held in conjunction with this year’s NIPS conference and DAC students presented during that event as well.

At the main conference, Sirui Yao presented “Beyond Parity: Fairness Objectives for Collaborative Filtering” (Yao and Bert Huang, assistant professor of computer science). An overview video for the paper can be viewed on YouTube.

DAC faculty Jia-Bin Huang, assistant professor of electrical and computer engineering collaborated on two papers which were also presented at the main NIPS conference: “Semi-Supervised Learning for Optical Flow with Generative Adversarial Networks” (with the University of California MERCED); and “MaskRNN: Instance Level Video Object Segmentation” (with the University of Illinois in Urbana-Champaign).

Yuliang Zou presented “Label-Efficient Learning of Transferable Representations across Domains and Tasks” (Zou collaborating with Stanford University and the University of California, Berkeley).

Both Yao and Zou received NIPS Travel Awards.

A Best Paper award went to “Co-trained Ensemble Models for Weakly Supervised Cyberbullying Detection” (Elaheh Raisi and Bert Huang), presented by Raisi during the conference workshop on Learning with Limited Labeled Data: Weak Supervision and Beyond.

“Detecting Adversarial Attacks on Neural Network Policies with Visual Foresight” (Jia-Bin Huang and Yen-Chen Lin, visiting student, collaborating with Nvidia Research and the National Tsing Hua University in Taiwan) was presented by Lin at the conference workshop on Machine Deception.

Raisi and Yao presented posters at the Women in Machine Learning Workshop. Raisi presented “A Weakly Supervised Deep Model for Cyberbullying Detection” (Elaheh Raisi, Bert Huang); and Yao presented “Fairness and Accuracy in Recommendation with Imbalanced Data Sparsity” (Sirui Yao, Bert Huang).