COMPLEX DATA ANALYTICS
Making sense of data is one of the great challenges of our century. While it is becoming easier to collect all kinds of data, from personal medical data, to online, scientific, public, and/or commercial data, creating interpretable knowledge from this data and separating out the consistent patterns from the noise is a very challenging task. Analyzing such data often requires developing new methodologies, models and computational tools. My main research interest is to think about these problems from theoretical, applied and computational angels. I love to collaborate with researchers from other disciplines and learn about their interesting problems and possibly develop novel methodologies and new statistical techniques to address their research problems.
Statistical Learning with Rank-Based Data
The cost of taking measurements from sampling units is an important factor in implementing a statistically sound sampling design in many applications. Sometimes researchers can use expertopinion knowledge or inexpensive measurements to obtain more representative samples from the population using rankbased sampling (RBS) designs. Straightforward applications of available learning methods in the literature on highly structured RBS data force stringent assumptions that might not be appropriate. A major emphasis of my past and current research program concerns developing novel rankbased statistical models, automatic learning algorithms and computational tools for various classification and regression problems using Bayesian and frequentist methods
In my lab, we develop novel rank-based predictive models, deep neural networks, support vector machines and finite mixture models for efficient prediction, classification, and clustering purposes. The ultimate goal is to actively embed the rank information into the learning process in order to either increase the efficiency of statistical inference or reduce the cost of study.
Rank-based support vector and deep learning machines for classification
Rank-based predictive models using Bayesian and frequentist methods
Finite Mixture modeling with RBS data
Resampling methods based on RBS data
Supervised and fractionally supervised learning with RBS data
Are you interested in
“Statistical learning Methods for Complex Data Analysis”?
Then send your application to join my research team!
Research Priorities at the moment
Despite the vast literature on RBS designs, there is not enough research pertinent to the theory and application of such designs in modern statistical learning with big and/or high-dimensional data. My current research program concerns developing new methodologies, learning algorithms and computational tools using observations obtained from RBS designs and revolves around the following long-term objectives:
- Developing New Statistical learning methodologies
- Developing Learning algorithms and computational tools
- Real world application and consulting services
New Methodologies
- Rank-Based SVMs for classification
- Quantile Regression with Nominated Samples
- Statistical Inference with Judgment Post stratified data
- Regression Studies based on Ranks
Learning algorithms
- Deep Learning for Ostoporosis diagnosis
- Random Forest for Sleep Apnea diagnosis
- Classification with selected or observed order statistics
- Logistic regression with nominated samples
- Fractionally supervised learning methods
Areas of application
- Medical Research (Osteoprosis diagnosis, Alzhimer disease, sleep apnea, …)
- Electrical Engineering
- Mechanical Engineering
- Biomedical Engineering
- Environmental Studies
- Quality control and Reliability analysis
Awards & Education
Winnipeg Rh Institute Foundation Award for outstanding contributions to scholarship and research in the natural science category
Rh Award
It is awarded to scholars who show “exceptional innovation, leadership, and promise” early in their careers
2012
Faculty of Science Award for Best Mentor
Faculty of Science
Established Researcher
2018
Faculty of Science Award for Research Excellence
Faculty of Science
Mid Career Researcher
2018
Merit Award for Research
University of Manitoba
2011, 2014