Data Scientist

*/*/**** ******* ******

https://resume.creddle.io/ 1/1

TAEHOON

ROH

*******-***@*******.***

905-***-****

taehoon-roh

throh313

**** *********** ****,

Oakville, ON, L6M3X2

Skills

DATA ANALYTIC TOOLS

Python

SQL

SAS

R

Excel

Matlab

BIG DATA TOOLS

Hadoop

Hive

Spark

VISUALIZATION

Tableau

PowerBI

Seaborn

AMAZON AWS

EC2

Redshift

S3

EMR

Client Projects

Data Science Specialist Nov. 2018 to Current

Woodbury Solution Toronto, ON

Client: e-Commerce startup

Currently working in an agile data science team to help an E-Commerce startup analyze data sets using AWS cloud, SQL, Python and Tableau Developed complex SQL queries to extract data from an analytics warehouse with 180 source tables

Developed merchant report, customer life time value model, and single customer view report using Python, SQL, and Tableau

Worked with marketing team to conduct targeted customer acquisition and reduced the reporting turnaround time by 80%

Worked in a agile team that runs weekly sprint planning sessions and daily stand-up meetings

Data Science Projects

KKBox's Customer Churn Prediction Jan. 2019

Build machine learning models to predict whether a user will churn after their membership expires

Prepared 21 GB raw data sets including user logs, transaction and user information using Python, Spark and SQL

Handled imbalanced data using over-sampling approach Trained machine learning models to predict churn

Tools : Python, Spark, SQL

Machine Learning Algorithms : Logistic Regression, RandomForest, AdaBoost, Gradient Boost, XGBoost

Bank Loan Default Nov. 2018

Build predictive machine learning models for loan defaults in banking data sets. Cleaned and merged 8 source tables including loan, order, transaction, account, etc. Prepared the loan-level modeling data sets using Python and trained machine learning models to classify loan default

The tuned model achieved AUC score of 0.93

Tools: MySQL, Python (SQLAlchemy, Pandas, Scikit-Learn) Machine Learning Algorithms: Logistic Regression, RandomForest, AdaBoost, Gradient Boost

Soccer Result Prediction Oct. 2018

Build machine learning models to predict soccer matches result Scraped soccer players market value data and soccer matches results Developed predictive machine learning models

The model achieved 71% accuracy score

Tools: Python (Beautiful Soup, Pandas, Scikit-Learn) Machine Learning Algorithms: Logistic Regression

Education

WeCloudData Data Science Bootcamp Oct. 2018 to Current York University June 2018 to July 2018

Certificate Big Data Analytic

York University Sept. 2012 to Apr. 2016

B.A. Spec. Hons. Mathematical For Commerce (Actuarial Stream)

Contact this candidate