Muntabir Choudhury

Old Dominion University

• Assisting and mentoring students with coursework, homework, and semester project for Computer Science courses including Machine Learning, Web Programming, and Computer Architecture.
• Assisting students with their semester project on search engine development for CS418/518 -- Web Programming. The main objective of this course is to educate students on the LAMP environment: Linux, Apache, MySQL, and PHP.
• Collaborating with Virginia Tech for the grant from IMLS on mining book length scanned documents such as Electronic These and Dissertations using machine learning and deep learning techniques.
• Publishing research work in top conference such as ACM/IEEE conference proceedings.

Bihrle Applied Research Inc

• Contributed to a project (i.e., BNSF Railway and FAA’s Pathfinder Program) to study and develop technologies for drone-based supplemental inspection of railway infrastructure.
• Developed and enhanced algorithms for train detection used by Rail-Inspector – a cloud-based software that process aerial imagery of railroad tracks using machine learning and deep learning.
• Built ground truth by labeling images for trains, used deep Learning model such as Fully Convolutional Network for segmentation, analyzed, and optimized the result.
• Implemented deep neural nets such as LeNET-CNN to train 82 classes of math symbols.
• Achieved an accuracy of 96% for detecting trains on railway and helped partial implementation of the segmentation model in the production cycle.

Los Alamos National Laboratory

• Researched on Offline Handwritten Mathematical Expression (HME) Recognition.
• Preprocessed math expressions image datasets such as rendering online images to offline, blurring, segmenting, and binary thresholding.
• Built a basic framework/pipeline for offline HME recognition and employed computer vision techniques for feature extraction -- contorur extraction and skeletonization.
• Implemented deep neural nets such as LeNET-CNN to train 82 classes of math symbols.
• Analyzed, visualized, and optimized the model to get the best performance model.

Resource9 Group, Inc.

• Analyzed captured snapshots by drilling down the full call stack and gaining code-level visibility to understand the root causes of the problem.
• Created health rules, policies, and alerts on the tier or node level to remediate any severe slowdown in the application.
• Identified the anomalies based on key performance metrics (KPIs) when threshold values exceeded.
• Used AppDynamics out of the box feature such as Development Level Monitoring.
• Configured POJO and POCO entry points for the business transactions which were not automatically detected.
• Identified hardware-level issues, such as memory leak detection, garbage collection, heap utilization, and thread contention.
• Followed AppDynamics best practices while educating the customers about using AppD special features such as Service End Point, Information Point (Code Metrics & Custom Metrics), Data Collectors (MIDC & HTTP).
• Integrated agents such as app agents, machine agents, analytics agent, DB agent, and EUM agent with the application.
• Analyzed log files such as Agent logs, BT logs, REST API logs, ByteCodeTransformation log for identifying the problems.

Elizabethtown College

• Researched relationship of companies and organizations, and analyzed their previous engagement with Elizabethtown College.
• Identified duplicate records using SQL, updated, and modified inconsistent data in the database.
• Optimized the relational data from the college database and analyzed it to predict the variable to understand philanthropy resources through various programs.
• Applied machine learning models such as Multiple Linear Regression and Backward Elimination to predict the variables.

• Used Jenzabar Ex for updating and creating relationships, salutations, and new records for potential donors and Alumni.
• Identified duplicate records to clean unnecessary data and avoid data inconsistency.
• Assisted pulling out records using SQL.
• Created new records for individuals or organizations who wanted to receive newsletter emails.

• Tutored students and assisted with coursework and assignments for Computer Organization and Architecture.
• Researched on PLC control (Nano LC Programmable Logic Controller and AXC PLCnext Control).
• Assisted students in the lab for Advanced Computer Engineering and graded over 20 assignments per week.