KARL TRUONG
ktruong.me@gmail.com | (+47) 47 73 95 29 | Trondheim, Norway______________________________________________________________________________________________________________________________PROFESSIONAL EXPERIENCE
NORWEGIAN UNIVERSITY OF SCIENCE AND TECHNOLOGY, NORWAY 06.2023 – Present
PhD Candidate
Be part of Advanced Analytics and Innovation team of the bank:
• Manage bancassurance project: performed statistical analyses and ML models to identify the most potential customers to offer insurance products (techniques: Logistics Regression, Gradient Boosting, Optuna); improving 40% of business conversion rate
• Build internal EDA and ML coding framework for the whole 30-data-specilist team on top of LightGBM, XGBoost, Scikit-learn, Optuna and Shap
TECHCOMBANK - GALAXY FINX DIGITAL, VIETNAM 04.2021 – 06.2023
Senior Data Scientist
Be part of Advanced Analytics and Innovation team of the bank:
• Manage bancassurance project: performed statistical analyses and ML models to identify the most potential customers to offer insurance products (techniques: Logistics Regression, Gradient Boosting, Optuna); improving 40% of business conversion rate
• Build internal EDA and ML coding framework for the whole 30-data-specilist team on top of LightGBM, XGBoost, Scikit-learn, Optuna and Shap
• Generate +600 features by PySpark for the Feature Store including demographics, products, transactions
• Communicate the analyses and findings to tech and non-tech high level management
HEINEKEN, VIETNAM 07.2020 – 04.2021
Data Scientist
Provide machine learning solutions for Sales, Trade and Brand departments:
• Identified the most +16,000 profitable business customers every quarter to run marketing campaign by random forest and logistics regression
• Assisted Business Controller to reduce potential +5M USD of cost by building an anomaly detection algorithm (k-nearest neighbors) to spot doubtful contract values
AIRBUS, FRANCE 09.2018 – 09.2019
Data Science Specialist – Apprenticeship
Used ML to standardize and shorten the aircraft manufacturing lead time:
• Contributed to research gap by developing three novel distance metrics on top of Euclidean distance
• Proposed 34 standard processes by unsupervised learning: DBSCAN, hierarchical and K-means clustering
• Measured assembly time of +3,000 aircraft components by solving underdetermined system of equation
VEOLIA - IRSTEA, FRANCE 04.2018 – 09.2018
Data Analyst Intern
Ensured input data quality in a project predicting water consumption in France:
• Imputed missing value in high dimensional data of +30,000 communes by R
• Performed statistical analysis: goodness of fit tests, mean t-test, quantiles, histogram, density plot, etc.
______________________________________________________________________________________________________________________________EDUCATION
NORWEGIAN UNIVERSITY OF SCIENCE AND TECHNOLOGY, NORWAY 06.2023 – Present
PhD in Statistics
Statistical Machine Learning to Detect Events of Interest in Distributed Acoustics Sensing
TOULOUSE SCHOOL OF ECONOMICS, FRANCE 08.2017 – 09.2019
Master of Statistics and Econometrics
Cumulative GPA: 15.5 / 20 (High Honors)
VIETNAM NATIONAL UNIVERSITY, VIETNAM 09.2012 – 05.2017
Bachelor of Finance and Banking
Cumulative GPA: 83.7 / 100 (High Honors)
CORVINUS UNIVERSITY OF BUDAPEST, HUNGARY 09.2015 – 06.2016
Erasmus Exchange Program
Cumulative GPA: 4.77 / 5 (Highest Honors)
______________________________________________________________________________________________________________________________
TECHNICAL SKILLS
Programming skills
• Python: NumPy, Pandas, Matplotlib, Seaborn, Scikit-learn, LightGBM, XGBoost, Optuna, Shap
• R: dplyr, tidyr, caret, mlr3, tseries, ggplot2, Shiny, ggmap
• Spark
• SQL
Data skills
• Machine learning:
- Supervised learning: gradient boosting, random forest, decision tree, logistics regression, linear regression, SVM, K-nearest neighbors
- Unsupervised learning: DBSCAN, Hierarchal clustering, K-means clustering
- General: hyperparameter optimization, probability calibration, model evaluation, explainable ML
• Feature engineering: imputation, handling outliner, categorical encoding, scaling, aggregating