Armita Kharmandar
Data Analyst
Armita Kharmandar recently graduated from the University of Victoria, BC, with a master’s degree in electrical and computer engineering with specialization of Machine Learning and Data Science. Armita joined the lab as a co-op student (data analyst) in summer 2024, and then continued her work in the lab post-graduation as a data analyst. Her current project involves working on an agent-based simulation study to examine the impact of different trajectories in the use of HIV pre-exposure prophylaxis to prevent HIV transmission, specifically among adolescent girls and young women in Eastern and Southern Africa.
In her role, Armita focuses on fine-tuning the model by identifying complex patterns and relationships within the extensive datasets generated by the agent-based model. Her responsibilities include designing and implementing a systematic set of model checks and refining the model by addressing demographic heterogeneity, analyzing fixed parameters, and conducting model parameterization and calibration to ensure accurate predictions of HIV transmission and intervention outcomes. Her forensic approach to model analyses and model checks are critical to ensuring transparency, reproducibility, and internal validity of the complex model.
During her master’s research in University of Victoria, she specialized in sentiment analysis, applying transfer learning and neural network models via the Keras API. Her background in model development, data mining, and pattern recognition equips her with the skills to utilize several AI-based models implementing different techniques of supervised, unsupervised, reinforcement learning and neural network, improve model accuracy and derive meaningful insights from complex datasets.
Looking ahead, Armita aims to broaden her expertise in mathematical modeling and the application of machine learning in public health to tackle global health challenges, with a particular interest in predictive modeling for health outcomes and interventions.
Mishra Lab 