The IACS Seminar Series will resume on Friday, February 17, 2017 with a talk by Dr. Sauro Succi, and IACS visiting lecturer.
IACS seminars are generally held every other Friday during the academic year, and are free and open to the public. Lunch is served at 12:30pm on a first-come, first served basis with the seminar beginning promptly at 1pm. Unless otherwise indicated, all seminars will be held in Maxwell Dworkin G115.
|February 17, 2017|
Boltzman and The Lattice: A Very Happy Computational Marriage
Speaker: Sauro Succi, IAC-CNR Rome, Italy and Harvard IACS
February 24, 2017
Connectomics is an area of neuroscience aimed to discover the neuron shape and interconnections (neural network) within animal brain. A comprehensive knowledge of the biological neural network is critically important for a complete understanding of brain functionality. Recent advances in Electron Microscopic (EM) imaging have enabled us to capture the minuscule cellular processes at nanometer scale. However, such ultra-high resolution recording also produces a massive amount of data that must be analyzed to extract the neurons. Practical neural reconstruction approaches rely on machine learning and computer vision algorithms such as segmentation and synaptic junction detection to perform this task. In this talk, I will describe how these tasks are addressed in multiple EM neural reconstruction efforts and mention a few biological discoveries made from some of these efforts.
March 3, 2017
Annual Dean's Lecture on Computational Science and Engineering.
Socially assistive robotics (SAR) is a new field of intelligent robotics that focuses on developing machines capable of assisting users through social rather than physical interaction. The robot’s physical embodiment is at the heart of SAR’s effectiveness, as it hinges on the inherently human tendency to engage with lifelike (but not necessarily human-like or otherwise biomimetic) agents. People readily ascribe intention, personality, and emotion to robots; SAR leverages this engagement stemming from non-contact social interaction involving speech, gesture, movement demonstration and imitation, and encouragement, to develop robots capable of monitoring, motivating, and sustaining user activities and improving human learning, training, performance and health outcomes. Human-robot interaction (HRI) for SAR is a growing multifaceted research area at the intersection of engineering, health sciences, neuroscience, social, and cognitive sciences. This talk will describe our research into embodiment, modeling and steering social dynamics, and long-term user adaptation for SAR. The research will be grounded in projects involving analysis of multi-modal activity data, modeling personality and engagement, formalizing social use of space and non-verbal communication, and personalizing the interaction with the user over a period of months, among others. The presented methods and algorithms will be validated on implemented SAR systems evaluated by human subject cohorts from a variety of user populations, including stroke patients, children with autism spectrum disorder, and elderly with Alzheimer's and other forms of dementia.
March 24, 2017Building Training Sets for Astronomical Data: A Bayesian Feature Transformation for Domain Adaptation
Supervised data mining and machine learning rely on the availability of labeled data. When sufficient training data is available, supervised models achieve high performance in many domains. However, labeled data is scarcer than unlabeled data and much more expensive and difficult to obtain. Moreover, when models that perform well in one setting are applied to data from a different but related domain -- e.g. from a different telescope or sensor -- performance often drops significantly. Additionally, the enormous rate at which unlabeled data is being generated in astronomy greatly surpasses the rate at which labeled data becomes available. Domain adaptation aims to learn from a domain where labeled data is available, the 'domain', and through some adaptation perform well on a different domain, the 'target domain'. In this talk, I will present a new probabilistic model that represents the source and target distributions as two Gaussian mixtures and finds a transformation between the feature spaces of the domains to transfer labeled data between them. Our approach allows working with data available in one domain as if it belonged to the other, enabling the training of models in the target domain from training sets adapted from the source domain. We evaluate our proposal in simulated data and the problem of variable star classification. In the latter, we use data from multiple different astronomical surveys with different characteristics in terms of sensor sensitivity, atmospheric conditions, and data sampling frequency, among others.
April 7, 2017
|April 14, 2017|
Speaker: Sudip Dosanjh, Lawrence Berkeley National Laboratory
Abstract: Will be posted shortly.