Kamol's Website

{# #} Kamol's Website

This portfolio showcases my academic work up until 2020. For my current professional endeavors, please contact me directly. Thank you!

Funded Projects

Disruption identification and dynamic mapping during disasters

Project name: ORDER-CRISP (funded by National Science Foundation (NSF))
Data used: Steaming Twitter data
Approach/Method: Document classification, Deep Learning, Geo-coding
Tools/Language Python, SKlearn, Keras

Find my publication here

Modeling the dynamics of hurricane evacuation

Project name: Disaster Analytics (funded by Southeast Transportation Center (UTC))
Data used: Steaming Twitter data, Demographic, and Hurricane Data
Approach/Method: Hidden Markov Model, Activity classification, Unsupervised/Semi-supervised Model
Tools/Language Python, SKlearn, ArcPy

Find my conference article here

Quantifying Human Mobility Resilience

Project name: Disaster Analytics (funded by Southeast Transportation Center (UTC))
Data used: Geo-tagged Twitter data
Approach/Method: Developed new approach to detect event and resilience
Tools/Language Python

Find my publication here

Sharing Real-Time Traffic Information With Travelers Using Twitter: An Analysis of Effectiveness and Information Content

Project name: Integrated Arterial Traffic Management (Funded by Florida Department of Transportation (FDOT))
Data used: Twitter data of FDOT official accounts
Approach/Method: Topic Modeling, Logit Model
Tools/Language Python, SAS, R, MS office

Find our publication here

Understanding the efficiency of social media based crisis communication during hurricane Sandy

Project name

(UTC)

Data used: Twitter Data
Approach/Method:Entory Analysis, Logit Model, Regression Model
Tools/Language Python, SAS, R

Find my publication here

Predicting hurricane evacuation traffic demand by integrating multiple data sources

Project name: Disaster Analytics (funded by Southeast Transportation Center (UTC))
Data used: Geo-tagged Twitter data, traffic sensor, and socio-demographic data
Approach/Method: Machine learning, LSTM Model
Tools/Language Python

Find our project reports here

An Agent-based Modeling Approach to Understand Human Behaviour and Infrastructure Disruption

Project name: ORDER-CRISP (funded by National Science Foundation (NSF))
Data used: Transportation network, Socio-demographic and Infrastruture Data
Approach/Method: Agent-based Simulation
Tools/Language Repast-Simphony, Java

Contact me for details

Understanding the influence of multiple information sources on risk perception dynamics and evacuation decisions: An agent-based modeling approach

Project name: ORDER-CRISP (funded by National Science Foundation (NSF))
Data used: Run-off data, Socio-demographic and Infrastruture Data
Approach/Method: Agent-based Simulation
Tools/Language Repast-Simphony, Java, Python

Contact me for details

COVID-19 Project and WebApp

Live Interactive Map to Visualize Spatio-temporal Comparison of COVID 19 Cases and Deaths

Project name: Personal Project
Data used: Covid 19 Data of Johns Hopkins University
Approach/Method: Web Development
Tools/Language Python, DJango, Folium, Amazon S3, Heroku

Take a look at the interactive map

Temporal Visualization of Covid 19 Appearance in Each County of the United States

Project name: Personal Project
Data used: New York Tiems Covid 19 data, County Shape file
Approach/Method: Web Based Visualization Using Python Folium
Tools/Language Python, HTML, JavaScript

Take a look at the GithubPage

High fidelity Location-based Risk Estimates for COVID-19 to Guide Reopening of Los Angeles

Project name: Los Angeles COVID 19 Data Science Challenge
Data used: Mobility Data, Covid 19 data.
Approach/Method: Agent Based Modeling
Tools/Language Repast Simphony, Java, Python, ArcGIS

Check out my github for details.

Challenge Website for Submission and Scoring

Project name: Personal
Data used: Annotated Twitter Data
Approach/Method: Web Design
Tools/Language Django, Python, Amazon S3

Check out the Challenge Website .

Collaboration Projects

Dynamics of social media communication in major hurricanes

Project name: Collaboration Project with N-HIRS (Florida International University)
Data used: Twitter Steaming data
Approach/Method: Dynamic Topic Model
Tools/Language Python, Gensim, ArcPy

Find my publication here

Modeling information diffusion and evacuation decision

Project name: Research Project
Data used: Twitter Steaming data, Evacuation Data
Approach/Method: Contagion Model, SIS/SIR model
Tools/Language Python

Find my INFORMS presentation abstract here

Individual/Challenge Projects

Modeling traffic operation in a complex traffic intersection using VISSIM simulation

Project name: Personal Project
Data used: Traffic flow, Signal, and Route choice data
Approach/Method: PTV VISSIM Traffic Simulation
Tools/Language VISSIM, R

Find the paper here

Uncovering the Impact of Climate Stress on Human Mobility and Resilience Using Mobile Phone Data

Project name: Data Challenge (Organized by Data for Climate Action (UN Global Pulse))
Data used: Mobile phone data, Socio-demographic Data.
Approach/Method: Event Detection Algorithm, Correlation Analysis, Lasso Regression
Tools/Language Python, ArcGIS

Contact me for details

Modeling bi-cycle crash using machine learning

Project name: (Transportation Data Challenge Hackathon)) (NSF Big Data Innovation Hub)
Data used: Capital Bike-share Data, Climate Data
Approach/Method: Machine Learning, Regression Analysis
Tools/Language Python, SKlearn, SPSS

Check out my github for details.