Navigation

O. DENIZ GENCAGA, Ph.D.

Alcoa Technology
Alcoa Technical Center

100 Technical Drive, Alcoa Center, PA, 15069, USA

d [dot] gencaga [at] ieee [dot] org

https://c3.ndc.nasa.gov/dashlink/members/45/

 

NEWS!!! I have been serving on the general Editorial Board of the Entropy Journal as of April 2013. I am also guest editing the special issue on “Transfer Entropy” of the same journal. You can reach the call for papers from the following page: http://www.mdpi.com/journal/entropy/special_issues/transfer_entropy

 

SUMMARY & RESEARCH INTERESTS:

I am interested in developing advanced mathematical tools to better understand various physical phenomena observed in diverse multidisciplinary areas, such as process control, manufacturing, control systems and intelligent data analysis.  My favorite research interests include the identification of relevant variables and their interaction in large, complex data sets using information-theoretic techniques and probabilistic graphical networks with causality and uncertainty analysis.

Some of my recent research topics include applications of machine learning, data mining and Bayesian signal processing methods to seismology, climatology and astronomy. I have contributed to the scientific and engineering communities by providing novel solutions to the following problems, which cannot be resolved satisfactorily using traditional approaches. Mostly, I took part in projects supported by NASA or international grants. These are outlined below and explained in detail in the “Projects” section:

  • Separation of natural and anthropogenic factors affecting surface temperature of Earth (2011-2012))
  • Applications of machine learning for the remote sensing of aerosols (2011-2012))
  • Machine learning techniques for missing data problems in remote sensing and space communications (2011-2012)
  • Assessment of climate feedback processes (2009-2011)
  • Identifying relationships among Earth climate data (2007-2009)
  • Characterization of interstellar organic molecules (2007-2009)
  • Sequential Bayesian modeling of non-stationary non-Gaussian processes and modeling seismic signals (2002-2007)
  • Synthetic Aperture Radar (SAR) image denoising (2005)
  • Adaptive signal processing: acoustic echo cancellation (1999-2000)

 

PROJECTS:    

1. Research Associate in the “Applications of machine learning for the remote sensing of aerosols”: I worked to detect aerosol optical depths (AODs) in the globe using satellite data and developing novel machine learning methods and comparing them with techniques such as the neural networks, support vector machines and Gaussian Processes.

2. As another project, I have been working on the separation of anthropogenic (man made) and natural influences on surface temperatures around the globe. Currently, I am developing Bayesian source separation techniques to analyze the problem. Generally, Bayesian source separation methods provide superior performances compared to their least squares counterparts, as the inversion (separation from the mixture) procedure is ill-posed.

3. Research Associate in “The assessment of climate feedback processes” funded by NASA

 Project description: Climate feedback processes play a very important role in general atmospheric dynamics. In order to better understand how Earth’s atmosphere behaves, I worked on understanding climate feedback processes, mostly focusing on cloud feedbacks. As the physical mechanisms involved in these processes are highly nonlinear and/or non-Gaussian, I utilized advanced information-theoretic statistical approaches that could handle such problems. These efforts mostly involved developing new Bayesian methodologies, which can help us model these physical processes by providing their uncertainties. Our research mostly focused on understanding which climate variables are related to particular tropical disturbances, such as the Madden-Julian oscillation (MJO), using large data sets obtained from more than a dozen satellite data product.

 4. Senior Research Support Specialist in “Identifying relationships among Earth climate data” funded by two NASA grants

 Project Description: In climate studies, identifying the relevant variables responsible for a phenomenon of interest is a highly challenging problem. To quantify the interactions between different relevant variables, we designed new Bayesian techniques where we estimate entropy and mutual information from data along with their error bars. Here, we significantly contributed to the field by providing the error bars of our estimations, as this is the most honest way of reporting one’s scientific results. We demonstrated these techniques on data taken from the International Satellite Cloud Climatology Project (ISCCP) and contributed to the field by providing successful ways of estimating mutual information from data, which is still a topic under discussion in the literature. 

 5. Senior Research Support Specialist in “Bayesian Source Separation for Astrophysical Spectra: Application to PAHs” funded by NASA AISRP

 Project Description: Complex organic molecules, known as Polycyclic Aromatic Hydrocarbons (PAHs), are widely observed in star-forming regions in interstellar space. To identify and classify PAHs, we proposed and applied new Bayesian source separation techniques to the infrared spectra recorded by the Infrared Space Observatory and the Spitzer Space Telescope. By our project, we introduced a new method to the community which outperforms the traditionally used Nonnegative Least Squares method as a useful alternative.

6. Research assistant in “Applications of Independent Component Analysis in Communications, Image Processing and Geophysics

Project Description: Earthquakes, tsunamis and hurricanes are examples of impulsive phenomena which may cause major disruptions. We developed a new Bayesian method to model time-varying impulsive signals and demonstrated this to model seismic data    taken from the Southern California Seismic Network Catalog. The probability distribution of the seismic signals are modeled by alpha-stable distributions. As we cannot express and solve for the equations of time-varying autoregressive alpha-stable distributions analytically, the sequential Monte Carlo based method that we proposed is a significant contribution for modeling signals possessing such nature. I also introduced the use of particle filtering to denoise Synthetic Aperture Radar (SAR) images to the community in one of my publications.

 PROFESSIONAL EXPERIENCE:

Dec. 2012- present:       Senior Scientist at Alcoa Technology, PA, USA.

Nov. 2011-Dec. 2012:    Research Associate at the William Hanson Centre for Space Sciences at the University of Texas at Dallas, Richardson,TX, U.S.A.

May 2011–Nov. 2011:    Postdoctoral research scientist at the Department of Physics and Astronomy, Middle Tennessee State University, Murfreesboro,TN, U.S.A.

Jan. 2009 – May 2011:  Research Associate in the National Oceanic and Atmospheric Administration and Cooperative Remote Sensing and Technology Centre (NOAA-CREST) at The City College of the City University of New York (CUNY), New York, New York, U.S.A.

May 2007- Jan. 2009:    Senior Research Support Specialist at the Department of Physics, University at Albany, State University of New York, Albany, New York, U.S.A.

2000 – 2007:                 Research and Teaching Assistant at the Department of Electrical and Electronic Engineering, Bogazici University.

09/2003, 02-07/2004      Research Assistant at Consiglio Nazionale delle Ricerche (CNR), Istituto di Scienza e Tecnologie dell’ Informazione “A. Faedo”,  Pisa, Italy

Summer 1996                Intern at NCR Corporation.

EDUCATION:                Ph.D., Bogazici University, Department of Electrical and Electronic Engineering,2007. Thesis Title: Sequential Bayesian Modeling of Non-Stationary Non-Gaussian Processes.

Thesis web page: http://theses.eurasip.org/theses/1/sequential-bayesian-modeling-of-non-stationary/download/

 AWARDS:

  •  Elevation to “Senior Member” status of IEEE, 2012.               
  •  NATO Research Fellowship in Italy, 2004;
  •  Best Student Paper Award, 13th IEEE Conference on Signal Processing and Communications Applications, May 2005.
  •  The European Research Consortium for Informatics and Mathematics (ERCIM) "Alain Bensoussan" Fellowship (I accepted another offer)

 PROFESSIONAL MEMBERSHIPS:      

1999-present The Institute of Electrical and Electronics Engineers (IEEE)

2012-present American Statistical Association (ASA)

2008, present    International Society for Bayesian Analysis (ISBA)

2012-present      Society for Industrial and Applied Mathematics (SIAM)

2013-present      Association for the Advancement of Artificial Intelligence (AAAI)

2012-present      American Society of Mechanical Engineers (ASME)

2013-present      Institute of Industrial Engineers (IIE)

2012-2013 Institute of Mathematical Statistics (IMS)

2012-present Association for Computing Machinery (ACM)

2011-2013         The American Association for the Advancement of Science (AAAS)

2010-2013       The New York Academy of Sciences (NYAS)

2005                 The European Association for Signal and Image Processing

2008-2012 American Geophysical Union  (AGU

2010-2013       American Meteorological Society (AMS)


 PROFESSIONAL ACTIVITIES:

2005    Reviewer for the 13th European Signal Processing Conference (EUSIPCO 2005), Sept. 2005.

2006    Reviewer for the 14th European Signal Processing Conference (EUSIPCO 2006), Florence, Italy, Sept. 2006.

2006    Reviewer for the 2006 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2006), Toulouse, France, May 2006.

2007    Reviewer for the 15th European Signal Processing Conference (EUSIPCO 2007), Poznań, Poland, Sept 2007.

2007    Reviewer for the 27th International Workshop on Bayesian Inference and Maximum Entropy Methods in Science and Engineering (MaxEnt 2007), Saratoga Springs NY, USA, July 2007.

2011    Reviewer and organizing assistant for the 31st International Workshop on Bayesian Inference and Maximum Entropy Methods in Science and Engineering (MaxEnt 2011), Waterloo, Canada, July 2011

Conference Organization:

Local organizer for the 27th International Workshop on Bayesian Inference and Maximum Entropy Methods in Science and Engineering (MaxEnt 2007), Saratoga Springs NY, USA, July 2007.

 Editorial:

Member of the Editorial Board of the Entropy journal.

Journal Reviewing

 IEEE Transactions on Signal Processing, IEEE Transactions on Image Processing, IEEE Transactions on Information Theory, IEEE Transactions on Neural Networks, IEE Proceedings-Vision, Image and Signal Processing, Digital Signal Processing, Entropy, Journal of the Acoustical Society of America

 WORKS IN PROGRESS:

1. D. Gencaga, W. B. Rossow, K. H. Knuth, 2012, “Recipes for the estimation of information flow in a dynamical system,” under preparation.

2.  D. Gencaga, D. J. Lary, 2012, “Information theoretic analysis for the bias correction of MODIS aerosol optical depth,” under preparation.

3. D. Gencaga, D. J. Lary, 2012, “A survey on the machine learning techniques in separating natural and anthropogenic influences on surface temperature using historical data,” under preparation.

 

PUBLICATIONS:

11. D. Gencaga, N. K. Malakar, D. J. Lary, "Survey on the Estimation of Mutual Information Methods as a Measure of Dependency versus Correlation Analysis," to be published in the Proc. of the 32nd Int. Workshop on Bayesian Inference and Maximum Entropy Methods in Science and Engineering (MaxEnt 2012), AIP Conference Proceedings, 2012.

2. N. K. Malakar, D. Gencaga, and D. J. Lary, “Towards Identification of Relevant Variables in the Aerosol Optical Depth Bias between MODIS and AERONET observations,” to be published in the Proc. of the MaxEnt 2012, AIP Conference Proceedings, 2012.

3. N. K. Malakar, D. J. Lary, A. Moore, D. Gencaga, B. Roscoe, A. Albayrak, M. Petrenko, and J. Wei, “Estimation and Bias Correction of Aerosol Abundance using Data-driven Machine Learning and Remote Sensing,” accepted for the NASA Conference on Intelligent Data Understanding: CIDU 2012, National Center for Atmospheric Research (NCAR), Boulder, Colorado, October 24 - 26, 2012.

4. D. Gencaga, E. E. Kuruoglu, A. Ertuzun, “Modeling non-Gaussian time-varying vector autoregressive processes by particle filtering”, Multidimensional Systems and Signal Processing, Vol. 21, Number 1, pp. 73-85, March 2010.

5. D. Gencaga, E. E. Kuruoglu, A. Ertuzun and S. Yildirim, “Estimation of Time Varying Autoregressive Symmetric Alpha Stable Processes using Gibbs Sampling”, Signal Processing, Vol. 88, Issue 10, pp. 2564-2572, October 2008.

6. D. Gencaga, A. Ertuzun, E.E. Kuruoglu, “Modeling of non-stationary autoregressive alpha-stable processes by particle filters”, Digital Signal Processing, Vol. 18, Issue 3, pp. 465-478, May 2008.

7. D. Gencaga, D. F. Carbon, K. H. Knuth, “Characterization of Interstellar Organic Molecules,” Proceedings of the 28th International Workshop on Bayesian Inference and Maximum Entropy Methods in Science and Engineering, AIP Conference Proceedings, Volume 1073, pp. 286-293 (2008).

8. K. H. Knuth, D. Gencaga, W. B. Rossow, “Information-Theoretic Methods for Identifying Relationships among Climate Variables”, Earth-Sun Systems Technology Conference (ESTC 2008), Adelphi MD, June 2008.

9. D. Gencaga, E. E. Kuruoglu, A. Ertuzun, “Estimation of Time-Varying Autoregressive Symmetric Alpha Stable Processes by Particle Filters”, Technical Report, 2006-TR-04, ISTI-CNR, 2006.

10. D. Gencaga, E.E. Kuruoglu and A. Ertuzun, “SAR Image Enhancement using Particle Filters”, ESA-EUSC 2005: Image Information Mining – Theory and Application to Earth Observation, Frascati, Italy, October 2005.     

11. D. Gencaga and A. Ertuzun, “On The Performance Comparison of Gradient Type Joint-Process Estimators in Adaptive Signal Processing”, 13th European Signal Processing Conference (EUSIPCO 2005), September 2005.

12. D. Gencaga, E.E. Kuruoglu and A. Ertuzun, “Estimation of Time-Varying Autoregressive Symmetric Alpha Stable Processes by Particle Filters”, 13th European Signal Processing Conference (EUSIPCO 2005), September 2005.      

13. D. Gencaga, E.E. Kuruoglu and A. Ertuzun, “Bayesian Separation of Non-Stationary Mixtures of Dependent Gaussian Sources”, 25th International Workshop on Bayesian Inference and Maximum Entropy Methods in Science and Engineering, AIP Conference Proceedings, Volume 803, pp. 257-265 (2005) .

14. D. Gencaga, E.E. Kuruoglu and A. Ertuzun, “Time-varying autoregressive parameter estimation of Cauchy processes by particle filters” , 13th IEEE Signal Processing and Communications Applications Conference (IEEE-SIU 2005), May 2005.  (IEEE Best Student Paper Award) 

ABSTRACTS:

1. D. Gencaga, K. H. Knuth, W. B. Rossow, “Comparison of Information-Theoretic methods to estimate the information flow in a dynamical system,” New York Workshop on Computer, Earth, and Space Sciences, NASA Goddard Institute for Space Studies, Feb. 24-25, 2011, New York, NY, USA.

2. D. Gencaga, W. B. Rossow, K. H. Knuth, “Recipes for the estimation of information-theoretic quantities to analyze the information flow between different variables”, NASA Conference on Intelligent Data Understanding: CIDU 2010, Mountain View, CA, USA, Oct. 5-6, 2010.

3. D. Gencaga, M. K. Tse, W. B. Rossow, K. Knuth, “Estimating Entropies and Mutual Information with Error Bars”, 9th World Conference of the Int. Soc. for Bayesian Analysis (ISBA 2008), Australia, 2008.

4. K. H. Knuth, D. Gencaga, D. F. Carbon, “Searching for Complex Organic Molecules in Space”, NASA Applied Information Systems Research Program Workshop, Adelphi MD, May 2008.

PRESENTATIONS:

1. D. Gencaga, “Identification of Relevant Climate Variables using Information-Theoretic Approaches”, October 31, 2008, NASA Jet Propulsion Laboratory, Pasadena, California, USA (Invited Talk)

2. D. Gencaga, “Alpha Stable Processes and their Application to Seismic Data Modeling”, November 9, 2007, Department of Physics, University at Albany, Albany, NY, USA (Invited Talk)

3. D. Gencaga, “Dependent Component Analysis of Non-Stationary Gaussian Signals”, October 11, 2005, Consiglio Nazionale delle Ricerche (CNR), Istituto di Scienza e Tecnologie dell’ Informazione “A. Faedo”, Pisa, Italy (Invited Talk)

4. D. Gencaga, “Identification of Relevant Climate Variables using Information-Theoretic Approaches”, September 23, 2011, Middle Tennessee State University, Department of Computational Sciences, Murfreesboro, Tennessee, USA (Invited Talk)

4. D. Gencaga, “Sequential Bayesian Modeling of Non-Stationary Non-Gaussian Processes”, January 23, 2007, Department of Electrical and Electronic Engineering, Bogazici University. (PhD Thesis Defense)

5. D. Gencaga, “Comparison of Information-Theoretic methods to estimate the information flow in a dynamical system,” New York Workshop on Computer, Earth, and Space Sciences, NASA Goddard Institute for Space Studies, Feb. 24-25, 2011, New York, NY, USA.

6. D. Gencaga, “Recipes for the estimation of information-theoretic quantities to analyze the information flow between different variables”, NASA Conference on Intelligent Data Understanding: CIDU 2010, Mountain View, CA, USA, Oct. 5-6, 2010.

7.  D.  Gencaga, “Estimating Entropies and Mutual Information with Error Bars”, 9th World Conference of the International Society for Bayesian Analysis (ISBA 2008), Hamilton Island, Australia, July 21-25, 2008.

8.  D.  Gencaga, “Characterization of Interstellar Organic Molecules”, 28th International Workshop on Bayesian Inference and Maximum Entropy Methods in Science and Engineering, (MaxEnt 2008), São Paulo, Brazil, July 6-11, 2008.

9. D. Gencaga, “SAR Image Enhancement using Particle Filters”, ESA-EUSC 2005: Image Information Mining – Theory and Application to Earth Observation, Frascati, Italy, October 2005.

10. D. Gencaga, “On The Performance Comparison of Gradient Type Joint-Process Estimators in Adaptive Signal Processing”, 13th European Signal Processing Conference (EUSIPCO 2005), September 2005.

11. D. Gencaga, “Estimation of Time-Varying Autoregressive Symmetric Alpha Stable Processes by Particle Filters”, 13th European Signal Processing Conference (EUSIPCO 2005), September 2005.     

12. D. Gencaga, “Bayesian Separation of Non-Stationary Mixtures of Dependent Gaussian Sources”, 25th International Workshop on Bayesian Inference and Maximum Entropy Methods in Science and Engineering (MaxEnt 2005), San Jose, CA, USA, August 2005.  

13. D. Gencaga, “Time-varying autoregressive parameter estimation of Cauchy processes by particle filters” , 13th IEEE Signal Processing and Communications Applications Conference (IEEE-SIU 2005), May 2005.

TEACHING EXPERIENCE

Guest Lecturer:

“Bayesian Data Analysis and Signal Processing (Prof. Kevin H. Knuth)”, Fall 2007 semester, Department of Physics, University at Albany.

“Computational Physics (Prof. Kevin H. Knuth)”, Spring 2008 semester, Department of Physics, University at Albany.

“Computational Physics”, Cotaught with Nabin Malakar, Spring 2009 semester, Department of Physics, University at Albany.

Teaching Assistant in the following courses at the Department of Electrical and Electronic Engineering, Bogazici University between 2000-2007:

Digital Signal Processing, Introduction to Information Theory, Signals and Systems, Communication Engineering, Introduction to Speech processing, Numerical methods in Electrical Engineering, Electrical Circuits, Electrical Circuits Laboratory, Communication Laboratory (Designed the lab experiments in LabView)

Graduate Student Mentoring: I have helped several graduate students at some stages of their research projects at the University at Albany.