Data Scientist Information Technology

LE NGUYEN NHUT TRUONG

Data Scientist

[ **************@*****.*** Ó +(84-979****** HCMC, Viet Nam

https://www.linkedin.com/in/lnntruong/ https://github.com/truongcntn2017 PROGRAMMING

LANGUAGES AND

FRAMEWORK

Scala: Professional working proficiency

Airflow: Professional working proficiency

Python: Professional working proficiency

Scikit-learn: Professional working

proficiency

Tensorflow: Limited working proficiency

Pytorch: Limited working proficiency

R: Limited working proficiency

SQL Server: Limited working proficiency

STRENGTHS

Hard-working Responsible Creative

Communication skills

LANGUAGES

English: Professional working proficiency

Vietnamese: Native or bilingual

proficiency

INTERESTS

Data Mining

Statistics

Natural language processing

Data Structure And Algorithms

Artificial Intelligence

Object - Oriented Programming

ACADEMIC

ACHIEVEMENTS

First prize Zalo AI Challenge Computer

Vision Task in Internal branch 2021

Top 27 challenge competitions in

University of Science in 2019, 2020

Participate ACM - ICPC at University of

Science in 2018, 2019, 2020

Semi final Entropy Competition about

Data Analysis at John Von Neumann 2019

Top 10 Round 2 Code Tour at VNG 2020

EXPERIENCE

Data Scientist

DataNest

November 2022 - August 2023 HCMC, Viet Nam

Improving Model Credit Score from 0.67 AUC to 0.71 AUC

Build model and deploy Gender, Age. Gender AUC 0.92, Age Group 0.83 Accuracy, 73MMonthlyActiveUsersofViettel.

Improving 0.7 AUC Model Credit Score by Device Target Encoding

Null rate Analytic Improve 1.3 AUC Score.

Data Scientist (Middle)

VNG Corporation, Zalo Group, R&D Team

April 2021 – March 2022 HCMC, Viet Nam

Research Embedding Technical, Auto ML, User Segmenation, Probability Data Structure, Parallel Computing and Snorkel.

EDA and ETL 70MMonthlyActiveUsersofZaloandmodelingfor Income prediction multiple classifiers. AUC - mu: 0.92.

Define problem and Prediction Income by location in all wards ( 10K) in Vietnam.

EDA and ETL 25MMonthlyActiveUsersofBaoMoiandmodeling using Keyword extraction (KeyBERT) and embedding keywords.

Based on keyword predict for User Interest (using GraphSAGE is one type of GNN).

URL embedding (using Character Embedding, Word Embedding).

Literature Review Problem and define problem. EDA and ETL 300M Monthly Active Anonymous Users for Prediction Age, Gender of Anonymous Users.

Literature Review Problem and define problem. Clustering User Anonymous (using HDBSCAN).

Research: Anchor method is able to explain any black box classifier, with two or more classes.

Associate Data Scientist (Junior)

VNG Corporation, ZION, ZaloPay, Risk Team

October 2020 – December 2020 HCMC, Viet Nam

Research CatBoost model, stacking methods, shap value model.

EDA and Modeling with million users.

Improving precious ( from 0.7 to 0.8) and recall ( from 0.2 to 0.3) for a detector of promotion abuser (In Processing Transaction State). Data Scientist Fresher

VNG Corporation, ZION, ZaloPay, Risk Team

July 2020 – October 2020 HCMC, Viet Nam

Module 1: Soft skills and presentation skills

Module 2: System and Network: Distributed system, network with spark, airflow

Module 3: Analytic: EDA transactions data with pandas profiling

(Domain knowledge e-commerce, banking, e-wallet)

Module 4: Modelling: Modeling for a detector of promotion abuser

(promotion transactions data, imbalanced data)

Research: Smote for imbalanced data

Research: Boots method for imbalanced data

EDUCATION

B.Sc. in Information Technology

Honor program

Faculty of Information Technology,

VNU - HCMUniversityofScience,

Viet Nam

2017 2021 GPA: 8.1/10

M.Sc. in Information Technology

Honor program

Faculty of Information Technology,

VNU - HCMUniversityofScience,

Viet Nam

2021 Present

EXPERIENCE

Web Developer

Pascalia Asia

June 2019 – September 2019 HCMC, Viet Nam

Creating web application using the iterative development process. Allow users to securely and easily login to your web application.

Build and secure a web server with Restful Developer-friendly API.

RD OCR with tesseract and spacy. (Framework of python for Computer Vision, Natural language processing).

Contact this candidate