Recent technological advances enable us to collect huge amounts of data from multiple sources; however, extracting meaningful information remains to be the main challenge. In complex problems, the structure we are looking for is often buried in the data. In those cases, in particular, looking at the data from different aspects; in other words, jointly analyzing data from multiple sources, i.e., data fusion (also called multi-block, multi-view or multi-set data analysis), increases the chances of capturing the hidden dynamics. For instance, in metabolomics, biological fluids are measured using a variety of analytical techniques such as Liquid Chromatography-Mass Spectrometry (LC-MS), Gas Chromatography-Mass Spectrometry (GC-MS) and Nuclear Magnetic Resonance (NMR) Spectroscopy with an ultimate goal of identifying chemicals related to certain conditions such as diseases. Data measured using different analytical methods are often complementary and their fusion enhances knowledge discovery.

This project focuses on developing mathematical models/algorithms for data fusion. We plan to apply the developed methods in different domains including sensory data analysis, social network analysis and metabolomics.

 PI: Evrim Acar, Faculty of Science, University of Copenhagen, Denmark

 COLLABORATORS:

• Rasmus Bro, Faculty of Science, University of Copenhagen, Denmark
• Mathias Nilsson, Faculty of Science, University of Copenhagen, Denmark
• Lars Ove Dragsted, Faculty of Science, University of Copenhagen, Denmark
• Michael Saunders, Stanford University, Stanford, CA
• Anne Tjønneland, Danish Cancer Society, Denmark
• Tormod Næs, Nofima, Norway
• Tamara G. Kolda, Sandia National Labs, Livermore, CA
• A. Taylan Cemgil, Bogazici University, Istanbul, Turkey
• Bulent Yener, Rensselaer Polytechnic Institute, Troy, NY

  SOFTWARE:

• The MATLAB CMTF Toolbox version 1.0

  DATA:

• Prototypical Experimental Coupled Data Sets

 RELATED PUBLICATIONS:

 Conferences/Workshops

• A Flexible Modeling Framework for Coupled Matrix and Tensor Factorizations (E. Acar, M. Nilsson, M. Saunders), EUSIPCO, 2014.
• Optimal Weight Learning for Coupled Tensor Factorization with Mixed Divergences (U. Simsekli, B. Ermis, A. T. Cemgil, E. Acar), EUSIPCO, 2013.
• Structure Revealing Data Fusion Model with Applications in Metabolomics (E. Acar, A. J. Lawaetz, M. A. Rasmussen, R. Bro), IEEE EMBC, pp. 6023-6026, 2013.
• Coupled Matrix Factorization with Sparse Factors to Identify Potential Biomarkers in Metabolomics (E. Acar, G. Gurdeniz, M. Rasmussen, D. Rago, L. Dragsted, R. Bro), ICDM Workshop on Biological Data Mining and its Applications in Healthcare, pp. 1-8, 2012.
• Link Prediction via Generalized Coupled Tensor Factorisation (B. Ermis, E. Acar, A. T. Cemgil), ECML/PKDD Workshop on Collective Learning and Inference on Structured Data, 2012.
• All-at-once Optimization for Coupled Matrix and Tensor Factorizations (E. Acar, T.G.Kolda and D.M. Dunlavy), KDD Workshop on Mining and Learning with Graphs, 2011.

 Journals

• Structure-Revealing Data Fusion (E. Acar, E. E. Papalexakis, G. Gurdeniz, M. A. Rasmussen, A. J. Lawaetz, M. Nilsson, R. Bro), BMC Bioinformatics, 15: 239, 2014.
• Link Prediction in Heterogeneous Data via Generalized Coupled Tensor Factorization (B. Ermis, E. Acar, A. T. Cemgil), Data Mining and Knowledge Discovery, December, 2013.
• Understanding Data Fusion Within the Framework of Coupled Matrix and Tensor Factorizations (E. Acar, M. A. Rasmussen, F. Savorani, T. Naes, R. Bro), Chemometrics and Intelligent Laboratory Systems,  129: 53-63, 2013.
• Coupled Matrix Factorization with Sparse Factors to Identify Potential Biomarkers in Metabolomics (E. Acar, G. Gurdeniz, M. Rasmussen, D. Rago, L. Dragsted, R. Bro), International Journal of Knowledge Discovery in Bioinformatics, 3:22-43, 2012.
• Coupled Analysis of in Vitro and Histology Tissue Samples to Quantify Structure-Function Relationship (E. Acar, G. E. Plopper, B. Yener), PLoS One , 7(3): e32227, 2012.

TALKS AT CONFERENCES/WORKSHOPS & SEMINARS:

• KU Leuven, Belgium, June 2, 2014.
• Arctic Analysis, Ilulissat, Greenland, March 10-14, 2014.
• ERCIM 2013, University of London, UK, December 15, 2013.
• Gipsa-Lab, Grenoble, France, November 8, 2013.
• Xerox Research Centre Europe, Grenoble, France, November 7, 2013.
• Kadir Has University, Istanbul, Turkey, October 23, 2013.
• Biomedical Engineering Society (BMES) Annual Meeting, Seattle, WA, September 28, 2013.
• IEEE EMBC, Osaka, Japan, July 7, 2013.
• ICME Colloquium, Stanford University, CA, June 3, 2013.
• SIAM Conference on Computational Science and Engineering (CSE), Boston, MA, Feb. 25-March 1, 2013.
• SPIE Medical Imaging: Digital Pathology Conference, Orlando, FL, Feb. 10, 2013.
• SISTA Workshop on Tensors and Data Analysis, KU Leuven, Belgium, December 13, 2012.
• ICDM Workshop on Biological Data Mining and its Applications in Healthcare, Brussels, Belgium, December 10, 2012.
• SIAM Conference on Applied Linear Algebra, Valencia, Spain, June 18-22, 2012.
• Three-way Methods in Chemistry and Psychology (TRICAP), Bruges, Belgium, June 2-7, 2012.
• Linear Algebra/Optimization Seminar , Stanford University, CA, May 10, 2012.
• Sandia National Labs, Livermore, CA, May 7, 2012.

  WORKSHOPS/MINISYMPOSIA:

• Two-part minisymposia on data fusion methods based on matrix and tensor factorizations at SIAM Conference on Computational Science and Engineering (CSE), Feb. 25-March 1, 2013, Boston, MA (Organizers: Evrim Acar, A. Taylan Cemgil, Rasmus Bro)

  ON-GOING LITERATURE SURVEYS:

• Data Fusion based on coupled matrix/tensor factorizations
• Metabolomics (with a special focus on cancer metabolomics)

This work is funded by the Danish Council for Independent Research | Technology and Production Sciences and Sapere Aude Program under the projects 11-116328 and 11-120947.

 Maintained by: Evrim Acar