[cs-talks] IVC Seminar, Natasha Jaques, Monday March 13, 1pm @ Hariri Seminar Room

Harrington, Jacob Walter jwharrin at bu.edu
Thu Mar 9 15:00:00 EST 2017


Affective Computing for Modeling Happiness, Stress, and Social interaction
Natasha Jaques, PhD Student, Massachusetts Institute of Technology
Monday March 13th, 1pm – 2pm

Abstract:
Affective Computing is the study of how computational methods like computer vision, signal processing, and machine learning can be used to model and predict emotions and other affective phenomena. In this talk I will describe several projects that demonstrate how to leverage recent advances in machine learning to more effectively predict affective states. For example, by training neural networks on short, one-minute slices of the facial expressions and body language of two people in a conversation, we can accurately predict whether they will bond with each other up to twenty minutes later. Or, using daily monitoring of a person’s smartphone logs and physiological sensor data, we can detect whether they will report feeling happy, healthy, and stressed. By building more sophisticated machine learning models that are able to account for individual differences, we can increase our prediction accuracy by more than 10%, leading to highly effective models that may be able to detect early warning signs of anxiety and depression.

Bio:
Natasha Jaques is a Ph.D. student in the Affective Computing group at the MIT Media Lab, where she studies how to use advanced machine learning and deep learning techniques to predict and interpret people’s psychological states. Her work has been featured in the MIT Technology Review, Boston Magazine, and on CBC radio. At the NIPS 2016 machine learning conference, her work was part of the project that won Best Demo, and she was awarded Best Paper at the Machine Learning for Health Care workshop. During her recent internship at Google Brain, she developed a novel method for applying deep reinforcement learning to recurrent neural networks. Natasha has a Master’s in Computer Science from the University of British Columbia, and graduated from the University of Regina with an Honours B.Sc. in Computer Science and a B.A. in Psychology. She has won numerous awards and fellowships, including the Robert Wood Johnson Foundation Wellbeing Fellowship, Microsoft Research Graduate Women’s Scholarship, UBC CS Merit Scholarship, and the S.E. Stewart Award in Arts.

Related Papers:
Jaques N., Taylor S., Nosakhare E., Sano A., Picard R.,"Multi-task Learning for Predicting Health, Stress, and Happiness", NIPS Workshop on Machine Learning for Healthcare, December 2016, Barcelona, Spain.
http://affect.media.mit.edu/pdfs/16.Jaques-Taylor-et-al-PredictingHealthStressHappiness.pdf

Jaques, N., McDuff, D., Kim, Y. K., and Picard, R. W. "Understanding and Predicting Bonding in Conversations Using Thin Slices of Facial Expressions and Body Language," In Proceedings of Intelligent Virtual Agents, California, USA, September 2016.
http://affect.media.mit.edu/pdfs/16.Jaques-IVAbonding.pdf

Jaques, N., Taylor, S., Azaria, A., Ghandeharioun, A., Sano, A., and Picard, R. "Predicting students' happiness from physiology, phone, mobility, and behavioral data" In Proc. Affective Computing and Intelligent Interaction (ACII), Xi'an, China, September 2015.
http://affect.media.mit.edu/pdfs/15.Jaques-Taylor-et-al-PredictingHappiness.pdf

Taylor, S., Jaques, N., Chen, W., Fedor, S., Sano, A., and Picard, R. "Automatic Identification of Artifacts in Electrodermal Activity Data" In Proc. International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), Milan, Italy, August 2015.
http://affect.media.mit.edu/pdfs/15.Taylor-Jaques-et-al-ArtifactDetectionEDA.pdf


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://cs-mailman.bu.edu/pipermail/cs-talks/attachments/20170309/82fed297/attachment.html>


More information about the cs-talks mailing list