Worked within the Data Science team of the LCSB on the prediction of symptoms of Parkinson's disease patients from wearable sensors data : • Benchmarked several methods on the BEAT-PD DREAM Challenge data, improving best methods to predict symptoms specifically for each patient, but also across different patients• Applied innovative solutions such as meta-learning algorithms• Compared the results to those obtained on the LuxPARK cohort

• Set up Hadoop-YARN and Spark on a cluster of CPUs• Created a segmentation algorithm mixing split and merge approach, andwatershed algorithm• Proposed an implementation to parallelize this solution on Spark