Simanta Barman

Graduate Student at the University of Minnesota

"Here is just one example of the total wrongness of something I tend to be automatically sure of: everything in my own immediate experience supports my deep belief that I am the absolute centre of the universe; the realest, most vivid and important person in existence. We rarely think about this sort of natural, basic self-centredness because it’s so socially repulsive. But it’s pretty much the same for all of us. It is our default setting, hard-wired into our boards at birth. Think about it: there is no experience you have had that you are not the absolute centre of. The world as you experience it is there in front of YOU or behind YOU, to the left or right of YOU, on YOUR TV or YOUR monitor. And so on. Other people’s thoughts and feelings have to be communicated to you somehow, but your own are so immediate, urgent, real." -David Foster Wallace

About



My Erdös number is 5:
Paul Erdös Menachem Magidor Sarit Kraus Peter Stone Michael Levin Simanta Barman

3

Peer-reviewed Journal Publications
Transportation Research Part C,
Transportation Research Record

4

Publications in Preparation
Optimal Ramp Metering
Optimal Airline Overbooking model
Variations of Max Pressure Control
Shared Autonomous Vehicle Dispatching

5

Sponsored Projects
National Science Foundation (NSF)
Minnesota Department of Transportation (MnDOT)
Minnesota Local Road Research Board (LRRB)
Kambr

7

Conference Presentations
INFORMS
Annual Meeting of the Transportation Research Board
American Society of Civil Engineers, ICTD
CTS Transportation Research Conference
International Symposium on Dynamic Traffic Assignment

Doctor of Philosophy in Computer Science0%
Master of Science in Industrial and Systems Engineering100%
Master of Science in Civil Engineering100%
Bachelor of Science in Civil Engineering100%

Resume

Summary

Extremely motivated learner and published researcher with experience using mathematical modelling, optimization and software design to solve engineering problems

Strong leadership, teamwork, and project management skills evidenced by 5 sponsored projects including 2 National Science Foundation sponsored projects

Excellent communication skills resulting in 3 peer-reviewed journal papers, 7 conference presentations and several seminar talks

Education

Doctor of Philosophy in Computer Science

Sept 2023 -

University of Minnesota, Twin Cities

Minneapolis, MN, USA

Master of Science in Industrial & Systems Engineering

Jan 2022 - May 2023

Focus: Analytics

University of Minnesota, Twin Cities

Minneapolis, MN, USA

Master of Science in Civil Engineering

Jan 2021 - Dec 2022

Focus: Transportation Engineering

University of Minnesota, Twin Cities

Minneapolis, MN, USA

Bachelor of Science in Civil Engineering

Feb 2015 - Apr 2019

Major: Transportation Engineering

Minor: Structural Engineering

Bangladesh University of Engineering and Technology (BUET)

Dhaka, Bangladesh

Work Experience

Graduate Research Assistant

Jan 2021 - Present
CAVe Lab, Department of CEGE, University of Minnesota

Minneapolis, MN

  • Analysis of throughput properties, optimal locations and performance of Max-pressure signal control
  • Traffic flow estimation for pedestrians and bicyclists using mobile data

Graduate Researcher

Sep 2022 - Dec 2022
Department of ISyE, University of Minnesota & Kambr

Minneapolis, MN

  • Optimal airline overbooking recommendation estimation based on demand forecasts using Machine Learning

Research Assistant

Nov 2019 – Jun 2020
Bureau of Research, Testing and Consultation, BUET

Dhaka, Bangladesh

  • Analyzing relationships between different pavement deflection measurements

Undergraduate Researcher

Mar 2018 - Apr 2019
Department of Civil Engineering, BUET

Dhaka, Bangladesh

  • Automatic traffic surveillance for local vehicles using transfer learning

Publications

1. Barman, S. & Levin, M. W. (2023). Throughput Properties and Optimal Locations for Limited Deployment of Max-Pressure Controls. Transportation Research Part C: Emerging Technologies, 150, 104105. Link

2. Barman, S., & Levin, M. W. (2022). Performance Evaluation of Modified Cyclic Max-Pressure Controlled Intersections in Realistic Corridors. Transportation Research Record, 03611981211072807. Link

3. Xu, T., Barman, S., Levin, M. W., Chen, R., & Li, T. (2022). Integrating public transit signal priority into max-pressure signal control: Methodology and simulation study on a downtown network. Transportation Research Part C: Emerging Technologies, 138, 103614. Link

Presentations/Posters

Barman, Simanta and Levin, Michael, Throughput Properties and Optimal Locations for Limited Deployment of Max-Pressure Controls.

Presented at 102nd Annual Meeting of the Transportation Research Board in Washington, D.C. (January 2023)

Barman S. & M. W. Levin. Performance evaluation of modified cyclic max pressure controlled intersections in realistic corridors.

Presented at:

  1. INFORMS in Anaheim, CA (October 2021)
  2. CTS Transportation Research Conference in Minneaplis, MN (November 2021)
  3. 101st Annual Meeting of the Transportation Research Board in Washington, D.C. (January 2022)
  4. ASCE ICTD 2022 in Seattle, WA (June 2022)

Barman S., M. W. Levin & R. Stern. Efficient Traffic Flow Estimation using Mobile Data Sources for Pedestrian and Bicycle Traffic Considering All Possible Paths.

Presented at:

  1. ASCE ICTD 2022 in Seattle, WA (June 2022)
  2. 9th International Symposium on Dynamic Traffic Assignment 2023 in Evanston, IL (July 2023)

Xu, T., S. Barman, M. W. Levin, R. Chen, & T. Li. Integrating public transit signal priority into max-pressure signal control: methodology and simulation study on a downtown network.

Presented at the 101st Annual Meeting of the Transportation Research Board, January 2022 in Washington, D.C.

Research Projects

TRAFFIC FLOW ESTIMATION USING MOBILE DATA

Aug 2021 – Current

  • Led a team of 3 and worked with 5 to develop a new method to estimate traffic flows
  • Formulated a non-linear optimization problem using the Markov property
  • Developed efficient solution algorithm using projected gradient descent in Python
  • Currently developing software solutions for the Minnesota Department of Transportation

CYCLIC MAX‑PRESSURE TRAFFIC SIGNAL CONTROL

Jan 2021 – Current

  • Developed novel signal control with proven maximum throughput
  • Used the TraCI API to implement within Simulation of Urban Mobility (SUMO)
  • ∼50% reduction in delays in calibrated model of 7 Hennepin County intersections

OPTIMAL AIRLINES OVERBOOKING ESTIMATION

Sept 2022 – Dec 2022

  • Developed a methodology to decide optimal overbooking number for an airline
  • Trained machine learning models including neural networks using airline datasets to forecast flight demands
  • Formulated and solved stochastic optimization problem to solve for optimal overbooking.
  • Worked in a team of 3 and reported to 4 others

LIMITED DEPLOYMENT OF MAX‑PRESSURE CONTROL

Jan 2021 – Aug 2022

  • Proved maximum throughput properties of Max-pressure control with limited budget
  • Formulated a mixed-integer-linear-program (MILP) to determine optimal location for signal control installation
  • Developed an efficient greedy solution algorithm to solve the MILP
  • Implemented solution in Python using Gurobi to find optimal location
  • Wrote code to create simulations in Java to validate results
  • ∼35% reduction in average travel times per vehicle with 100 intersection installs

MAX‑PRESSURE CONTROL WITH BUS RAPID TRANSIT PRIORITY

Apr 2021 – Mar 2022

  • Developed and proved maximum throughput property of Max-pressure control with bus rapid transit priority
  • Implemented the new signal control using Gurobi, TraCI API and SUMO in Python
  • Developed efficient solution algorithm using gradient descent in Python
  • ∼25% reduction in travel times for both bus and private vehicles in calibrated models
  • Collaborated with a team of 4

Miscellaneous Projects

Machine Learning Algorithms

    From Scratch using Numpy
  • K-means Clustering
  • Principal Component Analysis
  • Discriminant Analysis
  • Naive Bayes
  • Decision Tree
  • Multi-layer Perceptron / Neural Network
    Using PyTorch
  • Multi-layer Perceptron / Neural Networks
  • Convolutional Neural Networks (CNN)
  • Recurrent Neural Networks (RNN) for natural language processing: Includes Gated Recurrent Unit (GRU), Long Short Term Memory (LSTM)

Traffic Algorithms

  • Shortest Paths: Dijkstra's Algorithm, A-star
  • Time Dependent Shortest Paths: With and Without first-in-first-out (FIFO)

    • Static Traffic Assignment Algorithms
    • Link Based: Method of Succesive Averages (MSA), Frank-Wolfe, Conjugate Frank-Wolfe, Simplical Decomposition, Diagonalization
    • Stochastic User Equilibrium (SUE): STOCH algorithm using MSA
    • Path Based: Gradient Projection

Non-Linear Optimization

  • Implemented using non-linear optimization formulation:
    • Support Vector Machine (SVM), Linear Regression, Logistic Regression, Perceptron Learning
  • Algorithms:
    • Steepest Descent, Stochastic Subgradient Method, Coordinate Descent
  • Step Size Rules:
    • Constant, Diminishing & Armijo step size rule

Optimization

    Implemented Solvers:
  • Unconstrained Optimization
    • Gradient Descent using bisection & golden-section search
    • Newton Raphson

  • Simplex Algorithm
  • Simulated Annealing
  • Genetic Algorithm

Discrete Optimization

  • Knapsack using Dynamic Programming
  • Knapsack Cover Inequalities
  • Multiple Choice Knapsack
  • Graph Coloring

  • Elden Ring Armor Optimizer (Class Project)

Chess Deep Q Network

    AlphaZero like Chess Engine

    • Q-learning variations:
    • Tabular method for storing Q-values
    • Deep Q network
    • Deep Q network with experience replay

    • SARSA variations:
    • Tabular method
    • Deep Q network
    • Deep Q network with experience replay

Conways Game of Life

    Implemented in:

  • C++
  • Python

Contact Me

LinkedIn

GitHub

Email