Fast Clustering of Flow Cytometry Data via Adaptive Mean Shift
Date:
I was involved in the development of novel Machine Learning algorithms for the Clustering of Flow Cytometry data at the Life Sciences department in BD Biosciences. My work focused on improving the scalability of the existing Clustering framework at BD via Parallelization and/or adding approximations without affecting quality. Additionally it involved analyzing the data to come up with better distance metrics for characterization of the concept of ‘neighborhood’. Given the large volume and high dimensionality (>= 10) of the data, Adaptive Locality Sensitive Hashing (ALSH) was employed for efficient KNN search.