Data Analyst Manager

Kuo Song

Jersey City, NJ ***** +1-347-***-**** *********@*****.*** LinkedIn

EDUCATION

Fordham University New York, USA

Master of Science in Business Analytics, GPA: 3.9/4.0 Aug. 2019-Dec. 2020 Courses: Data Structure, Text Analytics, Web Analytics, Data Mining, Machine Learning, Database Management, Big Data Beijing Forestry University Beijing, China

Bachelor of Science in Finance and Banking, GPA:3.7/4.0 Sep. 2014-June 2018 TECHNICAL SKILLS

• Programming: Python (Scikit-learn, NumPy, Pandas), R (tidyverse, ggplot2), MySQL, Hadoop (Hive), Spark, SPSS

• Data Analytics tools: Tableau, Power BI, AWS, GCP, Google Analytics, Excel, QlikView, Spotfire, Alteryx

• Machine Learning: Logistic Regression, Random Forest, Boosting Tree, SVM, Clustering, KNN, Recommender System PROFESSIONAL EXPERIENCE

Alvarez & Marsal New York, USA

Data Analyst Intern June 2020-Sep. 2020

• Led a group of 6 to deliver three end-to-end data driven approaches to improve company revenue growth by 8%

• Developed churn model to predict customers’ behaviors and improve retention and decreased the churn rate by 2%

• Leveraged clustering analysis to identify the characteristics of high-value customers for target marketing strategies

• Identified cross-selling opportunities for 570K+ customers through market basket analysis to increase revenue per sale

• Visualized data insights in Tableau dashboards and presented actionable business recommendations to senior managers Fordham University, Science Center for Digital Transformation New York, USA Data Analyst Intern June 2020-Nov. 2020

Training trader robot using reinforcement learning

• Analyzed the seasonality trend of 200+ stock and performed correlation analysis to identify 4 key financial indicators

• Collected and transformed 10-year historical stock data; built a robot trader and trained is decision-making strategy on stock trading with neural network to maximize stock earning rewards, reached model accuracy of 83%

• Designed strategy with mixed-policy including exploration and exploitation, tuned weighting parameters and improved decision-making based on earnings derived from Moving Average Convergence Divergence (MACD) Empirical study on Green bond and Environmental, Social and Governance (ESG) performance

• Collected and transformed Green Bonds’ data since 2006, leveraged data manipulation packages in Python to ensure data integrity; loaded data into dashboard to perform exploring analysis, figured out the changing pattern of ESG disclosures

• Improved the model performance by conducting feature engineering, such as creating new features and one-hot encoding

• Built Logistic and Linear Discriminant Analysis model to identify the key factors leading to the increment of ESG scores, reaching the model accuracy of 82%

PROJECTS

Using Image Classification to Enhance Yelp’s Rating System

• Leveraged fine-tuned LeNet CNN model to quantify information of 200K images for 19K restaurants by predicting the probability of being good quality for each image

• Fed key variables derived from images and 34 other preprocessed variables into Random Forest models to boost Yelp’s rating performance, improved customers’ experience on restaurants with not enough reviews Retrieval Based Twitter Customer Service Chatbot

• Utilized natural language processing to clean, aggregate, and manipulate 3 million Twitter data of 15 global brands

• Extracted 10 topics from customer queries using topic modeling to better understand customers’ concerns with LDA

• Used 4 encoding algorithms to find most similar queries and instantly respond to customers, cutting human service by 60%

Contact this candidate