Türkiye
Worked on accelerating Deep Neural Network (DNN) Microservices. Found the performance bottleneck for different types of services (image classification and machine translation) on different platforms (CPU and GPU), analyzed latency-throughput trade-offs, and the behavior of the system.
Researched about missing data mechanisms, missing data handling methods, imputation techniques and their details. Implemented different missing data imputation techniques on Python. Performed several experiments with different datasets and different settings. Coded plotting and statistic test functions to interpret the results of the experiments.