User's Cover Picture
User's avatar

umer khalifa saleem raja

Healthcare Data Scientist

Message
Intro
Chennai, India
Data Scientist at Procyon Technostructure (Hekma AI)
Junior Research Fellow at Sathyabama Institute of Science and Technology
Studied Life Sciences at Sathyabama Institute of Science and Technology
Computer Software
linkedin.com/in/umer-khalifa-3152524a
Joined October 6, 2021

Skills

Languages

Tamil
-
Native or Bilingual
English
-
Fluent
About
I'm a data scientist with almost 5 years of experience in data science and machine learning. In the past two years, I'm exclusively working on multiple NLP projects.
Experience
User avatar
Procyon Technostructure (Hekma AI)
Dec 2019 – Present
Chennai and California
Data Scientist
Project: Activity Prediction (Completed) 1. Developed random forest regressor models for body activity prediction to correlate with heartbeat and sleep patterns for health monitoring. Project: Patient Identification for Clinical Trial (In Progress) 1. Developed deep learning-enabled orchestrated system to extract insights from trial eligibility Criteria to match patient information in EHR (Electronic Health Record). Project: Trial to Patient (In Progress) 1. Developing convolutional highway network for clinical criteria word embedding and taxonomy guided hierarchical embedding for EHR. 2. Similarly, attentional record alignment module with similarity loss term for matching trial criteria to patient EHR.
User avatar
Sathyabama Institute of Science and Technology
Jan 2016 – Present
Chennai
Junior Research Fellow
1. Predicted population dynamics for the year 2100 using machine learning algorithm and statistical package to visualize patterns and predict changes in response to climate change. 2. Prediction of prey-predator relationship and building model using R to determine the patterns of feeding against temperature and pH.
User avatar
Sathyabama Institute of Science and Technology
Jan 2018 – Dec 2019
Chennai
Senior Research Fellow
Project: Interactive influence of temperature, CO2, light, and nutrient fluctuations on marine plankton with implications to global climate change 1. Principle component and canonical correspondence analysis were used to figure out the response of multiple species to environmental variables. In addition, the Niche model was built to predict the position of the population in its Niche. 2. Furthermore, long-term satellite data of crucial environmental factors were extracted and analyzed using machine learning techniques to predict future changes in environmental parameters over time.
Education
User avatar
Sathyabama Institute of Science and Technology
Jul 2016 – Present
PhD, Life Sciences