1) Developed machine learning models to predict cardiac output in patients receiving mechanical circulatory support, utilizing in vivo high frequency time series device data obtained from pre-clinical and clinical experiments, achieving a 5.2% improvement over the previous algorithm approach.2) Conducted research and implemented a Domain-Adversarial Neural Network-based architecture to forecast aortic pressure in different patient cohorts, leveraging in silico models to generate high volume data.3) Analyzed and fine tuned the predicted probability values generated by a Deep Neural Network (DNN) model for predicting right heart failures, with the aim of enhancing the model's Brier skill score.

EECE2323 Lab for Fundamentals of Digital Design and Computer OrganizationProfessor :- Emad AboelelaAssisted in debugging Verliog code using Xilinx Vivado. Monitored and supported students for lab sessions.

Course:- EECE 4632 Hardware-Software Codesign for FPGA-Based SystemsProfessor :- Miriam LeeserAided students to implement Hardware/Software designs that involve using Python to write embedded processor code, and using FPGA fabric (PYNQ Board) to implement the hardware designs.

● Developed and implemented machine learning models with PyTorch to generate tags for an Elastic search-driven recommendation system. ● Constructed indexing and search scripts, leveraging the generated tags, and fine-tuned the scoring function for improved optimization.● Designed a Convolutional Neural Network (CNN) for vision-based tasks of identifying handwritten text and tables within images.● Developed a language-based Support Vector Machine (SVM) classifier to distinguish between different question subjects and to categorizequestions into either multiple-choice or other question types.● Created pipelines for preprocessing and storing client data and associated tags in databases like Redshift, S3, Dynamo, and Redis.● Examined performance challenges within the entire pipeline, eliminating redundant operations and fine-tuning the pipeline for a 15%reduction in response time.

•Designed and simulated gate driver circuits for IGBT and other Power device used in locomotives like SCR, Thyristor, power Mosfet etc. in Simulink.•Designed and simulated a PID controller for IGBT Gate driving Circuit.

•Developed a prototype module to generate test signals and interface this testing module to verify correctness of pulse oximeter calibration.