Loading...
 

Notable Events

Image

2017

April

New UROP, MEng and UAP projects are up! Visit our Student Project Page for details!

March

We welcome Dr.Stjepan Picek from KU Leuven, Belgium to his new position as post-doctoral associate with ALFA.

February

Congratulations to Dr. Erik Hemberg for his promotion to Research Scientist!

January

ALFA bids farewell to visitors Pasquale Salza of University of Salerno and Dr. Sarah al Hamoud of Saudi Arabia. We wish them the best of luck!
 

 

Image

 

2016

November 

Check up our updated information on 2016 UROPs/AUP/Pre-MEng and MEng projects here.

Congratulations to Dr. Erik Hemberg on being accepted into the 2015-2016 MIT Translational Fellows Program, an activity of the MIT Innovation Initiative.

ALFA's 2015 Summer and Fall publication and conference schedule is mostly determined. See ALFA's Publications

 

August

Bryce Kim won the National Science Foundation (NSF) Award for Young Professionals Contributing to Smart and Connected Health link

 

 

Image

2015

October 

Coverage in the New York Times and elsewhere of the STEALTH project.
Coverage in MIT News and elsewhere on the Data Science Machine.

 

August

Farewell to Dr. Ignacio Arnaldo and best wished for his new position at Pattern Ex. 

 

June

Farewell to Jacob Rosen and best wishes for his new job at MITRE. 

 

May

Jacob Rosen received Best Thesis Award from the Technology and Policy Program at MIT EDS department for his thesis "Computer Aided Tax Avoidance Policy Analysis".

Congratulations to Dr. Erik Hemberg on being accepted into the 2015-2016 MIT Translational Fellows Program, an activity of the MIT Innovation Initiative.ALFA's 2015 Summer and Fall publication and conference schedule is mostly determined. See ALFA's Publications page for the details.Coverage in the New York Times and elsewhere of the STEALTH project. Coverage in MIT News and elsewhere on the Data Science Machine.

Farewell to Dr. Ignacio Arnaldo and best wished for his new position at Pattern Ex.

Farewell to Jacob Rosen and best wishes for his new job at MITRE.Workshop on Data Privacy, SingaporeO'Reilly talks at NUS, NTU

Congratulations to Franck Dernoncourt on completing his S.M. Thesis.

Jacob Rosen received Best Thesis Award from the Technology and Policy Program at MIT EDS department for his thesis "Computer Aided Tax Avoidance Policy Analysis". 

Congratulations to ALFA's graduates: Bryan Collazo, Max Kanter, Alex Wang, Edwin Zhang (MEng,EEC) and Jacob Rosen (Master's TPP).

 

 

Image

2014

Una-May O'Reilly gave a talk entitled MOOCS: Research Collaboration, Data Privacy and the Role of Technology at an NSF Meeting on IRB, privacy and big data in Education.

Una-May O'Reilly gave a seminar entitled Formulating (big) Data Science Innovations for All as part of the speaker series at Dalhousie University's Institute for Big Data Analytics. At the CSAIL BigData Initiative's Fall meeting:

  1. Erik Hemberg and Jacob Rosen presented a poster entitled Concurrent Search of Tax Evasion Schemes and Tax Non-Compliance Indicators, co-authored with Una-May O'Reilly (ALFA) and Geoffrey Warner and Sanith Wijesinghe of MITRE.
  2. Franck Dernoncourt fielded a poster entitled Learning wavelet parameters using Gaussian process in massive signal data sets, co-authored by Kalyan Veeramachaneni and Una-May O'Reilly
  3. Kalyan Veermachaneni gave a presentation on Recommending Models for Big Data
  4. Una-May O'Reilly participated in the Panel on Data Privacy speaking on Privacy and Online Learner Data

ALFA research on MOOC student drop out prediction was highlighted with a guest post on the edX blog.

 

August 

ALFA welcomes a number of visitors. Sherif Halawa joins us from Stanford University. He is a member of the Lytics Lab. Varun Aggarwal, an alum of the group joins us from Aspiring Minds, India. Sylvain Cussat-Blanc joins us from University of Toulouse.


July 

Una-May O'Reilly and ALFA visitor Prof. Krzyztof Krawiec of won best paper in the Genetic Programming track at GECCO 2014. The paper, entitled "Behavioral Programming: A Broader and More Detailed Take on Semantic GP" will shortly be available via a free download link to the ACM digital library posted on our Publications page.


March 

Una-May, leader of ALFA, just returned from a two week trip to Asia. In Hong Kong she gave talks at both Hong Kong PolyU and Chinese UHK on ALFA's MoocDB and MoocVIZ projects.Una-May was accompanied by ALFA postdoc Erik Hemberg to Shantou University (STU), China on the same trip. The duo kicked off Week 1 of their new blended learning course. The course pilots a novel blended learning architecture. Five days of lectures and labs are face to face and on the STU campus during each of the first and final week of the course. Sandwiched in-between are 8 weeks of online learning via the OpenEDX platform. The course technically focuses on Evolutionary Processes and Computation with an eye to explicating the connection between evolutionary algorithms, artificial intelligence and evolutionary systems in action around us. It also emphasizes the importance of how learning takes place, encouraging reflection and introducing diverse learning activities including active learning, role play learning and peer teaching.


ALFA will present 3 papers at EuroGP in Granada, Spain in April. Post-doc Nacho Arnaldo and ALFA visitor Krzysztof Krawiec will attend. The three papers are listed on the ALFA publications page.


February

ALFA bids a fond farewell to our visitor Prof. Krzysztof Krawiec of Poznan University of Technology, Poland. Krzysztof co-published 4 papers with ALFA members during his Fullbright Scholars visit. For these publications, see the ALFA publications page.


ALFA's MOOC research team is discussed in a blog post authored by Dr. Lori Breslow. The post, entitled 'Working Across Disciplines to Improve Education appears in the Inside Higher Ed's BlogU.


January 

Dr. O'Reilly traveled to Taipei, Taiwan. At the Changing How the World Learns Symposium she presented a talk entitled Knowledge Mining Online Learning Data.

Dr. O'Reilly gave a talk, entitled Comparative Effectiveness using MIMIC II Clinical Data. It was presented at Critical Data: Secondary Use of Big Data from Critical Care.

 

 

Image

2013

December 

ALFA was featured on the blog of the Teaching and Learning Lab of MIT. See a post written by Jennifer DeBoer, entitled Facilitating Education Research in the MOOC Space.


Dr. O'Reilly gave an MIT XTALK Series lecture entitled MoocDB: Taming MOOC Big Data while Fostering Collaboration in Online Education Research.


Dr. O'Reilly gave an invited talk at the Machine Learning for Clinical Data Analysis and Healthcare - NIPS 2013 Workshop. It was entitled Secondary Use of Physiological Waveforms for Evidence-Generated Medicine.


November 

Dr. O'Reilly gave at talk entitled MoocDB: A Framework for Collaborative Online-Learning Behavioral Research at the BigData@CSAIL MIT Big Data Initiative at CSAIL Annual Member Meeting.


October

Dr. O'Reilly gave a talk entitled Knowledge Mining Blood Pressure Waveforms: The GigaBEATS Project at the Quantitative Medicine Series.

 

August

ALFA warmly welcomes Dr. Nacho (Ignacio) Arnaldo as a new post-doctoral associate. Nacho hails from the Parallel Architectures and Bioinspired Algorithms group at the Departamento de Arquitectura de Computadores y Automática of Universidad Complutense de Madrid . He brings expertise in approaching optimization problems with a combination of Evolutionary Computation and GPU computing.


ALFA warmly welcomes two Fullbright Award visitors. We will host Prof. Krzysztof Krawiec from Poznan University of Technology, Poland for 6 months. Prof. Conor Ryan from University of Limerick, Ireland will join us for 4.5 months.

 

July

Dr. O'Reilly visited Taipei for the 2013 MIT ILP-EPOCH Taiwan Symposium. She gave a talk on Scalable Machine Learning to Exploit Big Data for Knowledge Discovery. Presentation PDF

July 5-8, Dr. O'Reilly represented ALFA at the annual premier conference on evolutionary computation GECCO in Amsterdam, The Netherlands.

  • During the first part of the meeting, devoted to tutorials and workshops,
    • Dr. O'Reilly gave a tutorial Introduction to Genetic Programming. Powerpoint slides of this tutorial are available in the GECCO companion proceedings from the ACM Digital Library.
    • co-organized the enthusiastically received first meeting for Women@GECCO.
    • presented a paper entitled Efficient Training Set Use for Blood Pressure Prediction in a Large Scale Learning Classifier System at the International Workshop on Learning Classifier Systems, co-authored with ALFA members Erik Hemberg, Kalyan Veeramachaneni, Franck Dernoncourt and Mark Wagy.
    • gave an invited talk on FlexGP at the Symbolic Regression Workshop, co-authored with Kalyan Veeramachaneni, Owen Derby and Dylan Sherry, entitled FlexGP and Ensemble-Based Symbolic Regression.
  • Over the latter course of the conference, she
    • attended the editorial board meetings for Evolutionary Computation and Genetic Programming and Evolvable Machines and the ACM SigEVO executive board meeting.
    • judged the HUMIES competition with Professors Erik Goodman, Wolfgang Banzhaf and Darrell Whitley
    • presented a paper, co-authored with Kalyan Veeramachaneni and Owen Derby, entitled Learning Regression Ensembles with Genetic Programming at Scale in the Genetics-Based Machine Learning Track.
  • ALFA was also superbly represented by co-author Prof. Sylvain Cussat-Blanc, who presented a paper, co-authored by Dennis Wilson, E. Awa, Kalyan Veeramachaneni, and Una-May O'Reilly, entitled On Learning to Generate Wind Farm Layouts. This paper was the nomination from the Generative Developmental System Track for Best Paper Award.

ALFA Research Scientist Dr. Kalyan Veeramachaneni, Post-doctoral Associate Dr. Zach Pardos and graduate student Franck Dernoncourt attended the first annual MoocShop at Artificial Intelligence in Educational Data Mining. Dr. Veeramachaneni presented their co-authored paper entitled MOOCdb: Developing Data Standards for MOOC Data Science.

ALFA Research Scientist, Dr. Kalyan Veeramachaneni attended the 2013 International Conference on Machine Learning Workshop on Healthcare. The ALFA group paper, co-authored with Alexander Waldin and Una-May O'Reilly was entitled Learning Blood Pressure Behavior from Large Physiological Waveform Repositories.

 

June

Owen Derby was awarded his M.Eng and received second place in the MIT EECS department's Charles and Jennifer Johnson Master of Engineering Thesis Prize for Computer Science for his thesis entitled FlexGP: A Scalable System for Factored Learning in the Cloud

Dr. O'Reilly hosted a CSAIL Industry Affiliates Program virtual meeting on June 19. The topic was Scalable Systems for Large Scale Machine Learning. A recording of the meeting is available to IAP members on the CSAIL IAP website.

 

May

Dr's. Una-May O'Reilly and Erik Hemberg taught a week long, intensive course from May 6-10 on Evolutionary Processes and Systems at Shantou University, China.

Image

Dr. O'Reilly was a speaker at the Stanford Medical School's Big Data in BioMedicine Conference. Her talk, on May 22, 2013, was entitled The GigaBeats Project: Knowledge Discovery Methods for Blood Pressure Waveforms.Dr. O'Reilly was a speaker at the Spring CSAIL Industry Affiliates Program meeting. At the May 29th session on Applied Algorithms and Machine Learning she presented a talk entitled FlexGP: A Scalable System for Factored Learning on the Cloud

 

April

ALFA leader, Una-May O'Reilly was awarded the EvoStar Award for Outstanding Contribution of Evolutionary Computation in Europe. Una-May received the award at the 2013 EvoStar Conference reception at the glittering Vienna City Hall. Elena Marchiori also received the EvoStar award.

Image

Dr. Una-May O'Reilly addressed the 2013 Information and Communication Technologies Conference, hosted by MIT at the Tang Center, with a talk entitled Divide and Conquer: Machine Learning to Exploit Big Data Knowledge Discovery. A video recording of this talk is available from the MIT Industry Liaison Program. Date: April 24, 2013.

ALFA Group M.Eng student Owen Derby presented his co-authored paper at Evo-Applications, in the EvoPAR track. It is entitled Cloud Driven Island Model Genetic Programming by Owen Derby, Kalyan Veeramachaneni, Erik Hemberg, and Una-May O'Reilly. Follow our publications link on the sidebar for more information.

ALFA Group leader, Una-May O'Reilly presented a poster on a co-authored paper entitled Large Scale Island Model CMA-ES for High Dimensional Problems by Dennis Wilson, Kalyan Veeramachaneni, and Una-May O'Reilly also in the EvoPAR track of the Evo-Applications conference.

 

Image

 

2012

December

Una-May has returned from the Oxford-Stanford Conference on Big Data: Challenges and Opportunities for Human Health where she presented the EC-Star platform.

 


November

At Oxford University, group leader, Una-May O'Reilly addressed the Oxford-Stanford Conference on Big Data: Challenges and Opportunities for Human Health. Her  presentation described how the EC-Star platform would be of service to this community.


October

Group leader, Una-May O'Reilly gave a  presentation on the FlexGP project at the annual GE Whitney Symposium held at GE Global Research in Niskayuna, NY. The 2012 Symposium topic area was analytics, modeling, simulations, and decision-making in the age of industrial internet, big data, cloud, and supercomputing.

We've recruited David Xiao as a UROP to investigate a project in the context of our cloud-scale machine learning agenda. Another UROP, Ms.Diyang Tang will examine how well statistical machine learning methods for wind resource methods work on limited data collected from real wind farm sites.


We were honored to host Prof. Petros Drineas of Renssalaer (RPI). Prof. Drineas presented a talk entitled Randomized Algorithms in Linear Algebra and Applications in Data Analysis.

 

September

Dr. O'Reilly attended the VMWare Research Symposium.


EvoDesignOpt was admirably represented by new group member, Dr. Erik Hemberg, in the Presidential Inauguration Fall Beaver Dash where the race distance was 5k and a few smoots. Erik's time was 3rd in his class.

SUPER Super UROPS: Congratulations to Ulziibayar Otgonbaatar and Sebastian Leon for acceptance of their super UROP project proposals. Sebastian's project is entitled Mining data from MIT EdX. He will work with us to data mine 6.002X course data. Ulzi's project is entitled Building an Elasticity Module for Flex-GP. He will investigate extending our cloud-based version of FlexGP to elastically scale (i.e. contract and expand the number of learners simultaneously networked to work on a problem) in response to cloud compute prices (or price surrogates). Sebastian has been chosen to receive funding through the EECS Research and Innovation Scholars Program (RISP) and has been named an MIT EECS Undergraduate Research and Innovation Scholar. Ulzi has been will receive funding and has been named an MIT EECS - VMWare Undergraduate Research and Innovation Scholar. The Super UROP projects, with addition of other program requirements will support Ulzi and Sebastian in earning a Certificate of Advanced Undergraduate Research.


August 

New Research Opportunities With the Fall 2012 semester just around the corner we've identified a number of projects on different aspects of large scale machine learning and evolutionary algorithms. These projects are primarily aimed at MIT Juniors, Seniors, MEng, though keen sophomores can apply.

End of SummerThe summer is drawing to an end. We've welcomed Manushaqe (Manjola) Muco and Carah (Meijuan) Yan for short term (summer long) membership. Dennis Wilson joined us on a MIT Energy Initiative UROP while Owen Derby jump-started his M.Eng work on FlexGP. We will miss Mark Wagy and wish him the best at UVM in their Ph.D. program. In the Fall we welcome back Owen Derby and greet for the first time Dr. Erik Hemberg who will be a post-doctoral associate, Franck Dernoncourt who starts his EECS PhD and Monica Vitali who is visiting from Politecnico Milano. In November, Alex Waldin will visit from ETH Zurich. We congratulate Dr. Kalyan Veeramachaneni on his promotion to Research Scientist too!Don't forget to check out our updated research opportunities! They are intended for talented MIT students. Visitors to our group from beyond MIT are rare and typically require an established relationship.

 

April

Member of the Evolutionary Design and Optimization Group just won the best paper award in the EvoPAR track of the Evo-Applications conference. EvoPAR is devoted to Parallel Evolutionary Algorithms. The conference was held in Malaga, Spain, April 11-14, 2012. The paper is entitled FlexGP: Genetic programming on the cloud and is co-authored by Dylan Sherry, Kalyan Veeramachaneni, James McDermott and Una-May O'Reilly.

 

January 

MIT news published a spotlight article on our research on knowledge mining from sensory evaluation data. Read the full article here.The paper referred to in the spotlight article is Knowledge mining sensory evaluation data: genetic programming, statistical techniques and swarm optimization, authored by Kalyan Veeramachaneni, Ekaterina Vladislavleva and Una-May O'Reilly. The copy of record appears in Genetic Programming and Evolvable Machines here.

We have a new name! ALFA: Any Scale Learning for All is now the group formerly known as Evolutionary Design and Optimization. During the years our research agenda has extended beyond evolutionary computation approaches to design and optimization. The new name subsumes the old and adds emphasis to data or algorithms of any scale. It expresses our burgeoning interests in machine learning and MOOC learning. The "for all" highlights our commitment to application domains of social value such as healthcare, energy and education.