Seminars

Brains, Minds + Machines Seminar Series: Machine learning and AI for the sciences —towards understanding

Nov 17, 2017 - 2:00 pm
Klaus-Robert Müller
Venue:  MIT Building 46-3002 (Singleton Auditorium) Address:  43 Vassar St. Cambridge, MA 02139 Speaker/s:  Klaus-Robert Müller, Technische Universität Berlin

Abstract: In recent years, machine learning (ML) and artificial intelligence (AI) methods have begun to play a more and more enabling role in the sciences and in industry. In particular, the advent of large and/or complex data corpora has given rise to new technological challenges and possibilities. In his talk, Müller will touch upon the topic of ML applications in the sciences, in particular in neuroscience, medicine and physics. He will also discuss possibilities for extracting information from machine learning models to further our understanding by explaining nonlinear ML models.  E.g. Machine Learning Models for Quantum Chemistry can, by applying interpretable ML,  contribute to furthering chemical understanding.  Finally, Müller will briefly outline perspectives and limitations.

Organizer:  Joel Oller Organizer Email:  cbmm-contact@mit.edu

CBMM Special Seminar: Machines As Thought Partners

Sep 15, 2017 - 2:00 pm
Dr. David Ferrucci
Venue:  MIT Building 46-3002 (Singleton Auditorium) Address:  43 Vassar St. Cambridge, MA 02139 Speaker/s:  Dr. David Ferrucci

Abstract: AI systems should not only propose solutions or answers but also explain why they make sense. Statistical machine learning is a powerful tool for discovering patterns in data, but, Dr. Ferrucci asks, can it produce understanding or enable humans to justify and take reasoned responsibility for individual outcomes? 

Dr. Ferrucci will also include an overview of Elemental Cognition, his company that is focused on creating AI systems that autonomously learn from human language and interaction to become powerful and fluent thought partners that facilitate complex decision making. Specifically, Elemental Cognition investigates a future in which AI is a powerful amplifier of human creativity—a system that leverages statistical machine learning but focuses primarily on a type of learning that enables humans and machines to share an understanding and collaborate on exploring the question, “Why?” 

Speaker bio: Dr. David Ferrucci is the award-winning Artificial Intelligence researcher who started and led the IBM Watson team from its inception in 2006 to its celebrated success in 2011 when Watson defeated the greatest Jeopardy players of all time. Considered a landmark in AI, Watson delivered amazing results that out-performed all expectations. From 2011 through 2012, Dr. Ferrucci pioneered Watson's applications in health which helped lay the technical foundation for a new Healthcare Division at IBM.  In 2013, Dr. Ferrucci joined Bridgewater Associates, where he leads the Systematized Intelligence Lab. He joined to explore the application of AI in building explicable data-driven systems for optimal management and people analytics. 

Dr. Ferrucci’s more than 25 years in AI and his passion to see computers fluently think, learn, and communicate inspired him to found Elemental Cognition LLC in 2015. Elemental Cognition is focused on creating AI systems that autonomously learn from human language and interaction to become powerful and fluent thought partners facilitating complex decision making in areas ranging from healthcare to economics. AI system should not only predict possible solutions but should be able to explain why they make sense. 

Dr. Ferrucci graduated from Rensselaer Polytechnic Institute with a Ph.D. in Computer Science. He led numerous projects prior to Watson, and has over 50 patents and published papers in the areas of AI, Automated Reasoning, NLP, Intelligent Systems Architectures, Automatic Text Generation, and Automatic Question-Answering.  He was awarded the title of IBM Fellow and has won many awards for his work including the Chicago Mercantile Exchange’s Innovation Award and the AAAI Feigenbaum Prize.

Organizer:  Joel Oller Organizer Email:  cbmm-contact@mit.edu

Brains, Minds + Machines Seminar Series: Perceptual Organization From a Bayesian Point of View

Apr 21, 2017 - 4:00 pm
Venue:  MIT Singleton Auditorium, Room 46-3002 Address:  43 Vassar St. Bldg. 46 Cambride, MA 02139 Speaker/s:  Jacob Feldman (Rutgers)

Title: Perceptual Organization From a Bayesian Point of View

Abstract: Perceptual organization is the process by which the visual system groups the visual image into distinct clusters or units. In this talk I'll sketch a Bayesian approach to grouping, formulating it as an inverse inference problem in which the goal it to estimate the organization that best explains the observed configuration of visual elements. We frame the problem as an instance of mixture estimation, in which the image configuration is assumed to have been generated by a set of distinct data-generating components or sources (``objects''), whose structure, locations, and number we seek to estimate.  I'll show how the approach works in a variety of classic problems of perceptual organization, including clustering, contour integration, figure/ground estimation, shape representation, part decomposition, object detection, and shape similarity. Because the Bayesian framework unifies a diverse array of grouping rules under a single principle, namely maximization of the Bayesian posterior---or, equivalently, minimization of descriptive complexity---I'll argue that it provides a useful formalization of the somewhat vague Gestalt notion of Prägnanz (simplicity or "good form").

Joint work with Manish Singh, Erica Briscoe, Vicky Froyen, John Wilder and Seha Kim.

Organizer:  Guy Ben-Yosef Georgios Evangelopoulos Organizer Email:  gevang@mit.edu

Brains, Minds + Machines Seminar Series: The Convergence of Machine Learning and Artificial Intelligence Towards Enabling Autonomous Driving

Mar 24, 2017 - 4:30 pm
Photo of Prof. Amnon Shashua
Venue:  MIT Bldg. 10-250 Address:  77 Massachusetts Avenue, Cambridge MA 02139 Speaker/s:  Amnon Shashua - Hebrew University, Co-founder, CTO and Chairman of Mobileye

Abstract: The field of transportation is undergoing a seismic change with the coming introduction of autonomous driving. The technologies required to enable computer driven cars involves the latest cutting edge artificial intelligence algorithms along three major thrusts: Sensing, Planning and Mapping. I will describe the challenges and the kind of machine learning algorithms involved, and will do that through the perspective of Mobileye’s activity in this domain.

Biography: Prof. Amnon Shashua holds the Sachs chair in computer science at the Hebrew University of Jerusalem. His field of expertise is computer vision and machine learning. For his academic achievements, he received the MARR prize Honorable Mention in 2001, the Kaye innovation award in 2004, and the Landau award in exact sciences in 2005.

In 1999 Prof. Shashua co-founded Mobileye, an Israeli company developing a system-on-chip and computer vision algorithms for a driving assistance system, providing a full range of active safety features using a single camera. Today, approximately 10 million cars from 23 automobile manufacturers rely on Mobileye technology to make their vehicles safer to drive.  

In 2010 Prof. Shashua co-founded OrCam which harnesses the power of artificial vision to assist people who are visually impaired or blind. The OrCam MyEye device is unique in its ability to provide visual aid to hundreds of millions of people, through a discreet wearable platform. Within its wide-ranging scope of capabilities, OrCam's device can read most texts (both indoors and outdoors) and learn to recognize thousands of new items and faces.

Organizer:  Tomaso Poggio Organizer Email:  tp@ai.mit.edu

Brains, Minds + Machines Seminar Series: Towards machines that perceive and communicate

May 26, 2017 - 4:00 pm
Venue:  MIT Singleton Auditorium, Room 46-3002 Speaker/s:  Kevin Murphy (Google Research) Host: Josh Tenenbaum

Abstract: In this talk, I summarize some recent work in my group related to visual scene understanding and "grounded" language understanding. In particular, I discuss the following topics:

I will explain how each of these pieces can be combined to develop systems that can better understand images and words.

Bio: Kevin Murphy is a research scientist at Google in Mountain View, California, where he works on AI, machine learning, computer vision, and natural language understanding. Before joining Google in 2011, he was an associate professor (with tenure) of computer science and statistics at the University of British Columbia in Vancouver, Canada. Before starting at UBC in 2004, he was a postdoc at MIT.  Kevin got his BA from U. Cambridge, his MEng from U. Pennsylvania, and his PhD from UC Berkeley. He has published over 80 papers in refereed conferences and journals, as well as an 1100-page textbook called "Machine Learning: a Probabilistic Perspective" (MIT Press, 2012), which was awarded the 2013 DeGroot Prize for best book in the field of Statistical Science. Kevin is also the (co) Editor-in-Chief of JMLR (the Journal of Machine Learning Research).

Organizer:  Guy Ben-Yosef Georgios Evangelopoulos Organizer Email:  gby@mit.edu

Brains, Minds + Machines Seminar Series: What Can Machines Learn, and What Does It Mean for Occupations and Industries?

May 5, 2017 - 4:00 pm
Venue:  MIT, Bldg. 46-3189 Address:  43 Vassar St, Cambridge MA 02139 Speaker/s:  Erik Brynjolfsson (MIT)

Title: What Can Machines Learn, and What Does It Mean for Occupations and Industries?

Abstract: This talk will present a preliminary framework and approach for understanding the potential effects of machine learning (ML) on tasks, occupations and industries. Digital technologies have already had a substantial effect on the wages and income. The increased availability of high quality data and rapid advances in ML have the potential to generate even larger effects in the coming decade. The ultimate impact will depend in part on the feasibility, costs and capabilities of ML-based applications for various types tasks and the speed with which they are implemented. Workers, firms and industries with complementary investments (e.g. relevant skills, data, and technologies) are well positioned to benefit, while those whose tasks are easily substituted for by ML will likely face downward pressure on wages and prices. We are developing a taxonomy of tasks most suitable for ML and plan to estimate some implications of our model by analyzing data from a major online resume and job postings marketplace.

 

Speaker Bio: Erik Brynjolfsson is Director of the MIT Initiative on the Digital Economy, Professor at MIT Sloan School, and Research Associate at NBER. His research examines the effects of information technologies on business strategy, productivity and performance, digital commerce, and intangible assets. At MIT, he teaches courses on the Economics of Information and the Analytics Lab. Author or co-editor of several books including NYTimes best-seller The Second Machine Age: Work, Progress and Prosperity in a Time of Brilliant Technologies, Brynjolfsson is editor of SSRN’s Information System Network and has served on the editorial boards of numerous academic journals.

 

Organizer:  Georgios Evangelopoulos Tomaso Poggio Organizer Email:  gevang@mit.edu

CBMM Special Seminar: Machine Learning Techniques and Applications in Finance, Healthcare and Recommendation Systems

Apr 12, 2017 - 2:00 pm
David Vogel
Venue:  MIT Singleton Auditorium 46-3002 Address:  43 Vassar St, Cambridge MA 02139 Speaker/s:  David Vogel, Trustee (Voloridge Investment Management, LLC)

Abstract: The introductory portion of this talk will review some state-of-the-art machine learning techniques. We will discuss concepts of ensembles and popular methodologies within this category. We’ll touch upon collaborative filtering techniques used for recommendation systems, and we’ll present certain algorithms published specifically for healthcare models.

We will later focus on application of the mentioned machine learning techniques covering healthcare, recommendation systems and portfolio construction in finance. We will refer to some past data modeling competitions such as Netflix (2007) and the Heritage Health Prize (2014) where thousands of algorithms were pitted against each other and evaluated impartially on a withheld data set. Within the financial application we will present risk management models that can be used to dictate/constrain positions within the portfolio construction process.

Biography: David S. Vogel is an award-winning predictive modeling scientist. In 2009, he founded the Voloridge Investment Management, LLC and also serves as its Chief Scientist, Chief Executive Officer, Chief Technology Officer and Managing Member. He has earned international recognition for models ranging from medical applications to direct marketing and has won numerous modeling competitions. David has also been invited to speak at conferences and research institutes worldwide.

Organizer:  Tomaso Poggio Organizer Email:  tpoggio@mit.edu

CBMM Special Seminar: Spike Timing in Motor Control

Feb 21, 2017 - 4:00 pm
Ilya Nemenman
Venue:  MIT Singleton Auditorium, Room 46-3002 Address:  43 Vassar St, Cambridge MA 02139 Speaker/s:  Ilya Nemenman (Emory College)

Abstract: Last two-three decades have convinced the computational neuroscience community that sensory neurons encode information about the world not just in their firing rate, but also in the precise timing of their action potentials. However, whether this information is used by animals to actually drive behavior has never been shown. In this talk, in a time hopefully substantially shorter than a few decades, I will try to convince you that timing of spikes in multispike patterns, down to about 1 ms, is used by animals in their neural motor codes. I will focus primarily on control of vocalization and breathing in the Bengalese finch, a songbird, but will briefly discuss other animals too, from insects to monkeys.

This event is organized by the CBMM Trainee Leadership Council.

Organizer:  Wiktor Młynarski Organizer Email:  mlynar@mit.edu

On the forgetting of college academics and the role of learning engineering in building expertise

Feb 10, 2017 - 4:00 pm
Venue:  MIT Bldg. 46 Room 3002 Singleton Auditorium Address:  43 Vassar St, Cambridge MA 02139 Speaker/s:  Dr. Brian Subirana  Dr. Bror Saxberg

In this talk, Brian Subirana will first review research conducted over the last 120 years since Ebbinghaus's seminal work in 1885 on the forgetting curve. This review (joint work in progress with Aikaterini Bagiati and Sanjay Sarma) aims to understand whether, and to what extent, what is learned in the college classroom (if left unused) is likely to be mostly forgotten a “few” years after graduation: in general, does university knowledge and skills retention follows Ebbinghaus 1885 forgetting curve even decades after graduation? In the research reviewed so far there is no solid evidence that “unused” memories are retained.

Bror Saxberg will then review research on expertise and learning, and draw out implications and models that have been empirically tested.  He will then talk about how this work can influence practical learning engineering at scale, with examples from Kaplan. The talk will finish with bold speculations by both authors on implications of the findings presented for the future of University education and leave time to invite views and reactions from the audience.

Organizer:  Georgios Evangelopoulos

CBMM Special Seminar: Development of Cortical Representations in Human and Macaque Infants

Dec 2, 2016 - 4:00 pm
Rebecca Saxe and Margaret Livingstone
Venue:  McGovern Institute for Brain Research, MIT Bldg. 46 Room 3002 (Singleton) Address:  43 Vassar St, Cambridge MA, 02139 Speaker/s:  Margaret Livingstone and Michael Arcaro (Harvard Medical School) Rebecca Saxe (MIT BCS) Organizer:  Elisa Pompeo Organizer Email:  epompeo@mit.edu

Pages

Subscribe to Seminars