MINH ANH PHAM
I'm a data science engineer at FPT Artificial Intelligence in partnership with MILA, researching machine learning methodologies to build models focused on processing visual-based data and object recognition. I'm also in charge of instructing Fresher Data Science courses at FPT, which covers Big data processing and Database. Prior, I worked at Viettel Corporation - the largest technology & telecommunications corporation in Vietnam, pioneers to develop digital finance and data services.
My research interest lies in machine learning & computer vision with applications to computational photography and image processing. I believe by a better understanding of different aspects of AI, we can further scientific breakthroughs in not only economics but other fundamental issues facing society.
Research projects
Customer Score
The model will score accounts on a scale of 1000 and classify accounts according to thresholds of high risk, medium risk, and low risk are equivalent to the probability that virtual account of the business/point of sale.
Credit Scoring
A credit scoring model is a risk management tool that assesses the credit worthiness of a loan applicant by estimating her probability of default based on historical data. It uses numerical tools to rank order cases using data integrated into a single value that attempts to measure risk or credit worthiness.
Fraud Detection
Fraud detection to prevent money from being obtained through fake virtual mobile phone
Work experience
FPT Software®, Artificial Intelligent Department
Data Science Engineer
Nov 2021 - Present
- Research on machine learning methodologies to build models focused on identifying user interactions in the user interface on television for recommendations using PyTorch, YOLOv5, and ONNX.
- Lecturing Fresher Data Science Courses (15 students per class) Including: SQL For Data Science Course, Scala Programming, Fundamentals of Apache Hadoop and Apache Spark.
VIETTEL Digital®, Advanced Technology Center
Data Science Engineer
Mar 2020 - Nov 2021
- Build machine learning models utilizing telecom and digital transactional data for credit scoring using LightGBM and K-Folds.
- Lead 4-member team to build a Centralized Operating System that collects and analyzes data from Hadoop Distributed File Systems (HDFS), in which capable of processing real-time data.
- Design and implement data flow architecture practically working on highly constrained systems using Pentaho, Apache Spark, Apache Kafka, Apache Airflow, and Hadoop.
Let’s Talk Online - English Learning Social Network
Project Manager - FPT University, Capstone Project
Sep 2019 - Feb 2020
- The project based on the domains of WebRTC and SNRS (social network-based recommendation sys- tem), enhancing user recommendation engine by enriching the user rating matrix with the transitive associations among users.
- Lead a team of 5 members to successfully deliver the project and achieve Excellent Capstone Project Reward at the FPT Univerity 2019-2020
FPT Software®, JIRA Atlassian Technology Tools Department
Java Backend Developer
Jun 2018 - Sep 2019
- Provide technical solutions for optimizing system architecture that balanced the need to reduce redundancy with performance requirements.
- In charge of redesigning and delivering optimal Database Architecture.
Accomplishments
The Best Student of Software Engineering Programme
Vietnamese Outstanding Female Students in Science and Technology Award
Professor Nguyen Van Dao National Scholarship for talented students
100% of undergraduate program tuition fees, FPT University.
Vietnam National Petroleum Corporation scholarships for outstanding students
Recent Blog
Optimal Transport - Introduction
Applications/ Methods of Optimal Transport (OT): Brief Introduction
Credit Scoring Models
Some techniques in building credit score models.
Customer Scoring Models
Evaluate different machine learning models in building customer score.
Get in Touch!
This website showcases my research interest, past and current work, and experiences.