My thesis work focuses on machine learning and pattern recognition. I have designed algorithms that learns the underlying cluster dependent dissimilarity measure while seeking compact clusters. More specifically, I worked on automatically learning the Gaussian kernel in unsupervised way. I also explored the semi-supervised machine learning paradigms. I developed a clustering algorithm that uses partial supervision to guide the algorithm in learning a good partition of the data and to learn a local Mahalanobis distance measure with respect to each cluster. Furthermore, I applied my machine learning skills in the context of hyperspectral image analysis. I used the gradual assessment of the membership of elements in a set in a endmember detection framework. I developed an approach that attempts to estimate the mixing parameters of multiple models simultaneously.

I have acquired extensive presentation skills over the MRL weekly presentations. I built effective communication skills while strengthening the ability to structure presentations and handle question and answer sessions. I enjoy teaching, and I am always looking forward to opportunities to do so. I acted as instructor during my engineering job by lecturing on STMicroelectronics IPs to new recruited employee. I also supervised Engineering internship on medical and image processing at Medical image processing Laboratory of the medical university of Monastir, Tunisia. Besides, I lectured software languages and supervised computer science projects at "Jeune et science de Tunis" club, Tunisia.