On April 2nd, 2024, UCLA Graduate Division hosted its annual GradSlam competition. This year, a Bioengineering graduate student, Soulaimane Bentaleb, was one of the finalists for the competition and won 2nd place overall! Congratulations Soulaimane on this great achievement!
Read below about Soulaimane’s research work and his experience participating in GradSlam!
Research title and abstract
Research Thesis Title: Unsupervised Classification of Interictal iEEG Activity as Pathological or Physiological in Epilepsy Patients
Grad Slam Pitch Title: Automated and Personalized Detection of Epileptic Activity in Resting State
Abstract: With timely and proper treatment, an estimated 70% of the 65 million people with epilepsy worldwide could be seizure-free. Accurate detection of pathological/epileptic interictal activity can be used to treat the 30-40% of epilepsy patients who do not respond to anti-seizure drugs. However, interictal detection algorithms rely on supervised learning, which requires labeled training data and human intervention for patient-specific tuning. An unsupervised approach overcomes the latter by requiring no labels or intervention from medical staff. A fully unsupervised model was developed to accurately classify intracranial electroencephalography (iEEG) interictal activity as physiological or pathological. Three sets of features were extracted to yield 178 features: 27 statistics from the time domain, 7 statistics from the frequency domain, and 144 statistics from the time-frequency domain. Four unsupervised dimensionality reduction techniques were then combined with four unsupervised classification methods. Singular Value Decomposition (SVD) combined with a K-Means model achieved the best performance: a 90.6% average F-2 score and 1.82 F-2 score variance.
Name of faculty mentor
Wentai Liu
Q: What was your experience like participating in Grad Slam?
A: Scientific research and public speaking are two of my passions, so I very much enjoyed my experience participating in Grad Slam. It was a unique chance to share my work with a wide audience who would have never heard about it otherwise.
Q: How did you prepare to enter Grad Slam?
A: I came up with an initial presentation and pitched it to my friends. They provided great feedback and I iterated based on that. I then repeated this process of asking for recommendations and improving my presentation between every round of the competition, editing my pitch using the feedback provided by the judges and my friends.
Q: What is your advice to graduate students who would like to enter Grad Slam in the future?
A: Grad Slam is an amazing and unique opportunity we are very lucky to have. Make sure you take advantage of it! This was the highlight and culmination of my graduate time at UCLA, the best possible way to showcase my research work. That being said, I recommend participating in Grad Slam every year of your graduate degree as there is no requirement or expectation to have finished your research project to participate.
If you would like to watch the final competition, it is available on YouTube: https://www.youtube.com/watch?v=rUOBiED4H38