About Me
Hi, my name’s Sifat Moon and I am an applied computational scientist and an ex-software engineer with a decade of experience in research and development. My background ranges from machine learning, graph theory, network science, data mining, and predictive modeling. I have pioneered several novel computational tools to model and analyze graph-based spreading processes. In my Ph.D. and postdoc, I became an expert in developing end-to-end solutions leveraging real-world data.
I am most skilled in: Machine Learning, NetworkScience, Scalable Data Mining, High Performance Computing
Professional Experience
Network Systems Science and Advanced Computing division, Biocomplexity Institute, University of Virginia
Postdoctoral Researcher
July 2021 - Present
Charlottesville, VA, USA
-
Lead a project with an interdisciplinary team of three members to deliver bi-weekly update on a detailed knowledge graph project to find the economic impact of under-vaccinated spatiotemporal clusters.
– 3TB high-dimensional APCD (All-Payer Claims database) healthcare data, ICD-10 code, data mining, foundation model, graph neural network, recurrent neural network.
-
Develop an abductive agent-based surveillance tool to understand the political opinion on a social contact network with millions of nodes in a distributed environment.
– HPC (High-Performance Computing) system, pattern recognition, support vector machines, XGboost.
-
Build a graph neural network-based machine learning surrogate model for the agent-based simulation.
– Guide two junior scientists and more than four grad students.
Network Science and Engineering group, Kansas State University
Graduate Research Assistant
Aug 2016 – May 2021
Manhattan, KS, USA
- Designed a computationally efficient stochastic Monte Carlo simulation tool to understand the time-series local dynamics of the Markov spreading processes over large networks.
- Modified and implemented approximate Bayesian computation based on sequential Monte Carlo (ABC SMC) sampling method for a stochastic individual-level multi-layer network system.
- Directed a project to estimate a movement network from NASS data using a maximum entropy reinforcement learning method to support data privacy.
- Optimized watercolor brush algorithm and pattern fill tools for the s-pen drawing bitmap engine of Galaxy Note5.
- Developed and deployed Allshare play/Samsung Link software for Windows Phone 8.1 and an initial version of quick-connect (a convergence app of Samsung).
- Led a group of one software engineer and two interns for four months.
Education
Kansas State University
Ph.D. in Computer Engineering
2016 - 2021
CGPA: 3.82/4.00
Dissertation: Modeling and analysis of spreading processes over large networks from limited data.
Relevant Courese: Scalability for data science, multivariate statistical methods, network theory, machine learning and pattern recognition, analysis of algorithms, agent-based game theory, applied probability theory, and random process.
A Little More About Me
Alongside my interests in networks and software engineering some of my other interests and hobbies are:
- Reading
- Hiking