Select a calendar:
Filter November Events by Event Type:
SUNMONTUEWEDTHUFRISAT
Events for November 30, 2016
-
MHI CommNetS Seminar
Wed, Nov 30, 2016 @ 02:00 AM - 03:00 PM
Ming Hsieh Department of Electrical and Computer Engineering
Conferences, Lectures, & Seminars
Speaker: Mohammad Rasouli, University of Michigan
Talk Title: Capacity and Energy Markets for Stable Renewable Economy
Series: CommNetS
Abstract: Existing supportive mechanisms for investment on renewable energies are not sustainable with higher penetration level of renewables. Rather, these mechanism should be replaced by market mechanisms. On the other hand, spot markets in place for conventional energies already suffer from underinvestment problem and there is an ongoing debate on the use of capacity markets for motivating sufficient investment.
In this talk we discuss the economical origins of underinvestment in conventional economies, and how renewables will change the situation. We propose a block investment market mechanism with forward moving approach that has the following features. (F1) The expansion and production allocations corresponding to the unique Nash Equilibrium (NE) of the game induced by the mechanism are the same as those that maximize the sum of utilities of the producers and the demand. (F2) It is budget balanced. (F3) It is individually rational. (F4) It is price efficient that is, the price for electricity at equilibrium is equal to the marginal utility of the demand and to the marginal cost of production by producers with free capacity.
Biography: Mohammad Rasouli is a PhD student in EECS: Systems joint with MSc in Economics at University of Michigan. He has received his Bachelor in EE from Sharif University of Technology. He uses stochastic control, game theory and mechanism design to study emerging cyber-physical systems including energy systems and cyber-security.
Host: Prof. Ashutosh Nayyar
Location: Hughes Aircraft Electrical Engineering Center (EEB) - 248
Audiences: Everyone Is Invited
Contact: Annie Yu
-
PhD Defense - Greg Harris
Wed, Nov 30, 2016 @ 10:30 AM - 12:30 PM
Thomas Lord Department of Computer Science
University Calendar
Customized Data Mining Objective Functions
Ph.D. candidate: Greg Harris
Wednesday, Nov. 30, 2016
10:30 AM, EEB 110
Abstract:
Interpretable machine learning models, such as classification rule lists, enable knowledge discovery and model vetting by domain experts. Their transparency, however, often comes at the cost of accuracy, when compared to more complex models. Our research seeks to improve the accuracy of such models while retaining their interpretable rule-based form. Our strategy is to generate domain-dependent objective functions that specify heuristic trade-offs tailored for individual datasets.
Our first contribution is FrontierMiner, a new rule-based algorithm for predicting a target class with high precision. It learns a non-parametric objective function directly from the data. We show that FrontierMiner finds higher-precision rules more often than competing rule induction systems in a study involving 1,000 synthetic datasets and 138 real-world classification tasks. Our second contribution is PRIMER, a new algorithm for maximizing event impact on time series. It has an objective function that adapts to the level of noise in the data. It also incorporates user-provided input on the expected response pattern as a heuristic that helps prevent over-fitting. We show PRIMER is competitive with state-of-the-art regression techniques in a large financial event study, yet has improved model interpretability. Our third contribution is a method of learning an objective function from user feedback in the form of pairwise rankings. With this feedback, we use learning-to-rank algorithms to combine existing measures into an overall objective function that more closely matches the user's preference. We conclude the presentation with directions for future research.
Biography:
Greg Harris is currently a PhD candidate in the Computer Science Department at the University of Southern California. His research interests include data mining, pattern recognition, and machine learning. He also holds a Master of Financial Mathematics degree from the University of Minnesota and a Bachelor of Science degree in Applied Physics from Brigham Young University.
Defense Committee: Viktor Prasanna (chair), Cauligi Raghavendra, Ellis Horowitz
Location: Hughes Aircraft Electrical Engineering Center (EEB) - 110
Audiences: Everyone Is Invited
Contact: Lizsl De Leon
-
Astani Civil and Environmental Engineering Seminar
Wed, Nov 30, 2016 @ 12:00 PM - 01:00 PM
Sonny Astani Department of Civil and Environmental Engineering
Conferences, Lectures, & Seminars
Speaker: Mari Winkler, Ph.D., University of Washington
Talk Title: Do More in Less: Intensifying Wastewater Treatment Plants with Aerobic Granular Sludge Technology
Abstract: See Attachment
More Information: Environmental Engineering Seminar.docx
Location: John Stauffer Science Lecture Hall (SLH) - 100
Audiences: Everyone Is Invited
Contact: Evangeline Reyes
-
MHI/EE-Electrophysics Seminar, Wednesday, November 30th at 2PM in EEB 132
Wed, Nov 30, 2016 @ 02:00 PM - 03:30 PM
Ming Hsieh Department of Electrical and Computer Engineering
Conferences, Lectures, & Seminars
Speaker: David Allstot, University of California at Berkeley
Talk Title: Switched-Capacitor Circuits Research in Wireless Networks and Ultra-low-power Sensor Interfaces
Abstract: Emerging wireless standards aggregate information by selecting combinations of contiguous or non-contiguous channels, thereby enabling wider transmission bandwidths, and hence, higher data rates. Frequency-interleaved analog-to-digital conversion (FI-ADC) is an attractive emerging technique for carrier aggregation receivers because it facilitates an efficient way to dynamically vary the receiver bandwidth in order to address the many possible channel combinations. Compared to their time-interleaved counterparts, the specifications of the samplers in each parallel channel in FI-ADCs are significantly relaxed, thereby resulting in lower overall power consumption in the receiver. This work extends the FI-ADC idea to the quadrature frequency-interleaved oversampled data converter (QFI-ADC) to achieve greater aggregate data rates. Previously, digital-to-analog converter (DAC) and other inter-channel mismatches have limited the performance of QFI-ADCs. We propose a low-complexity element rotation algorithm (ERA) to mitigate DAC mismatches. The ERA is synthesized from the corresponding mismatch transfer function using a rigorous mathematical procedure which is shown to be generally applicable to low-pass, high-pass, band-pass and quadrature ERA's. Simulations confirm that the resulting low-complexity quadrature ERAs have advantages over previously proposed approaches in terms of both performance and hardware complexity. An additional gain calibration technique alleviates image folding due to mismatches between the quadrature DAC elements, which yields higher SNDR.
The original switched-capacitor power amplifier is a polar power amplifier, amplifying a non-constant envelope modulation by linear combination of the amplitude modulation and phase modulation. It has since been extended to operate across multiple supply domains and to operate using quadrature and multiphase signals. The proposed research will be important to extend the SCPA architecture for future applications. For instance, the SCPA can be used as an enabling technology for massive MIMO, where moderate power, highly efficient, versatile transmitter cores with moderate die area are needed. In massive MIMO, hundreds of transmitter chains drive hundreds of antenna elements to form communications beams that enhance data service to multiple individual users. In addition to massive MIMO, the SCPA offers high efficiency for low-power, high order modulation schemes that are being deployed for wireless sensor systems prevalent in the internet of things. Additionally the SCPA can offer enhanced out-of-band rejection by implementing digital filtering directly in the RF front-end circuitry of a transmitter. This can be by means of weighted summation of an array of small SCPAs. Finally the frequency range of the SCPA can be extended.
Finally, research on the use of CMOS ring amplifier circuits in bio-medical and other ultra-low-power sensor networks is discussed. An analog compressed sensing front-end is used to motivate further investigations.
Biography: David J. Allstot received the B.S., M.S., and Ph.D. degrees from the Univ. of Portland, Oregon State Univ., and the Univ. of California, Berkeley.
He has held several industrial and academic positions. He was the Boeing-Egtvedt Chair Professor of Engineering at the Univ. of Washington from 1999 to 2012 and Chair of the Dept. of Electrical Engineering from 2004 to 2007. In 2012 he was a Visiting Professor of Electrical Engineering at Stanford University and from 2013 to 2016, he held a three-year appointment as the MacKay Professor in Residence in the EECS Dept. at UC Berkeley.
Dr. Allstot has advised about 65 M.S. and 40 Ph.D. graduates, published more than 300 papers, and received several awards for outstanding teaching and research including the 1980 IEEE W.R.G. Baker Award, 1995 and 2010 IEEE Circuits and Systems Society (CASS) Darlington Award, 1998 IEEE International Solid-State Circuits Conference (ISSCC) Beatrice Winner Award, 2004 IEEE CASS Charles A. Desoer Technical Achievement Award, 2005 Semiconductor Research Corp. Aristotle Award, 2008 Semiconductor Industries Assoc. University Research Award, 2011 IEEE CASS Mac Van Valkenburg Award, and 2015 IEEE Trans. on Biomedical Circuits and Systems Best Paper Award. He has been very active in service to the IEEE Circuits and Systems and Solid-State Circuits Societies throughout his career.
Host: MHI/EE-Electrophysics
Location: Hughes Aircraft Electrical Engineering Center (EEB) - 132
Audiences: Everyone Is Invited
Contact: Marilyn Poplawski
-
MHI CommNetS Seminar
Wed, Nov 30, 2016 @ 02:00 PM - 03:00 PM
Ming Hsieh Department of Electrical and Computer Engineering
Conferences, Lectures, & Seminars
Speaker: Mohammad Rasouli, University of Michigan
Talk Title: Capacity and Energy Markets for Stable Renewable Economy
Series: CommNetS
Abstract: Existing supportive mechanisms for investment on renewable energies are not sustainable with higher penetration level of renewables. Rather, these mechanism should be replaced by market mechanisms. On the other hand, spot markets in place for conventional energies already suffer from underinvestment problem and there is an ongoing debate on the use of capacity markets for motivating sufficient investment.
In this talk we discuss the economical origins of underinvestment in conventional economies, and how renewables will change the situation. We propose a block investment market mechanism with forward moving approach that has the following features. (F1) The expansion and production allocations corresponding to the unique Nash Equilibrium (NE) of the game induced by the mechanism are the same as those that maximize the sum of utilities of the producers and the demand. (F2) It is budget balanced. (F3) It is individually rational. (F4) It is price efï¬cient that is, the price for electricity at equilibrium is equal to the marginal utility of the demand and to the marginal cost of production by producers with free capacity.
Biography: Mohammad Rasouli is a PhD student in EECS: Systems joint with MSc in Economics at University of Michigan. He has received his Bachelor in EE from Sharif University of Technology. He uses stochastic control, game theory and mechanism design to study emerging cyber-physical systems including energy systems and cyber-security.
Host: Prof. Ashutosh Nayyar
Location: Hughes Aircraft Electrical Engineering Center (EEB) - 248
Audiences: Everyone Is Invited
Contact: Annie Yu
-
CS Colloquium: Grace Hui Yang (Georgetown University) - Statistical Modeling of Information Seeking
Wed, Nov 30, 2016 @ 02:30 PM - 03:30 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Grace Hui Yang , Georgetown University
Talk Title: Statistical Modeling of Information Seeking
Series: CS Colloquium
Abstract: Many modern IR systems and data exhibit new characteristics which are largely ignored by conventional techniques. What is missing is an ability for the model to change over time and be responsive to stimulus. Documents, relevance, users and tasks all exhibit dynamic behavior that is captured in big data sets (typically collected over long time spans) and models need to respond to these changes. This talk provides an up-to-date introduction to statistical modeling of information seeking. In particular, I will talk about how we model information seeking as a partially observable Markov decision process and achieve high accuracy in the TREC Session Tracks. I will also talk about evaluation in dynamic information retrieval modeling and the TREC Dynamic Domain Track.
Biography: Grace Hui Yang is an Assistant Professor in the Department of Computer Science at Georgetown University. Grace obtained her Ph.D. from the Language Technologies Institute, Carnegie Mellon University in 2011. Grace's current research interests include dynamic search, search engine evaluation, privacy-preserving information retrieval, and information organization. Prior to this, she conducted research on question answering, ontology construction, near-duplicate detection, multimedia information retrieval and opinion and sentiment detection. Grace is a recipient of the National Science Foundation (NSF) Faculty Early Career Development Program (CAREER) Award. Grace co-chaired the SIGIR 2013-2014 Doctoral Consortium, SIGIR 2017 Workshop, and WSDM 2017 Workshop. She served as an area chair for SIGIR 2014-2017 and ACL 2016. Grace also co-organized the TREC Dynamic Domain Track since 2015.
Host: Cyrus Shahabi
Location: Olin Hall of Engineering (OHE) - 100C
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
