Graduate students interested in data mining, machine learning, data science applications or health informatics are welcome to apply, Please visit our Graduate Program and apply through our online system and send me a message. Postdoc: If you have secured independent funding covering all expenses, you have good references, and there is overlap between your research interests and my research, contact me.
I have been at the University of Alberta for more than 20 years. I specialize in Data Mining and Machine Learning.
I graduated more than 80 students in the last 20 years.
My list of publications gives the big picture of my research activities.
One of my research interests is associative classifiers. Associative classification is a supervised machine learning paradigm that hinges on association rule mining to learn classification rules, a representation that is transparent and human interpretable.
I am also working on learning with class imbalance, how to effectively oversample your rare classes or do cost-sensitive learning so your learned model is not overshadowed by the majority class.
After inference, it is often necessary to explain a decision or prediction made. Explainable AI is one of my research foci. We developed a framework that provides a form of explanation for a decision made by any blackbox ML.
Finally, I concentrate on Social Network Analysis, particularly what is known as Community Mining (finding densely interconnected sets of nodes) and unobserved Link Prediction, and this in typical graphs, in attributed graphs and probabilistic graphs (graphs where edges are not observed with certainty ).
I also work on applied Machine Learning in applications like Precision Health such as:
Definitely, I also use my techniques for the analysis of Social Media like Sentiment Analysis and lately False Claim Detection.
Finally, in Natural Language Processing I also work on and I am excited by Automatic Response Generation for Conversational Agents.
|
Here is my academic background:
|
