Jessie Le
I am currently pursuing MS Degree in Business Analytics with Concentration in Statistics at Drexel University, Philadelphia.
My specialization are statistical and econometric modeling, data mining, optimization algorithms, machine learning and natural language processing applied to finance, consulting, healthcare and manufacturing.
Portfolio
Data Analytics Projects
Electric Vehicle Sale Querying and Reporting Project
Technical skills: SQL, SAP, Tableau
This is my project in Information System Course: Database Analysis and Design for Business Analytics . I built executive report to answer key business questions in sale of electric vehicles. There are three seperate data sets provided in this course: Demographic, State Population and Electric Vehicle Sale. I used SAP HANA to join and extract data from these data sets and used Tableau to visualize answers for these business questions.
MLS Player Salary Analysis
Technical skills: SAS, Clustering, Statistical Tests
This is my team project in MIS633: Predictive Business Analytics with Relational Database . This project was inspired by a huge fan of football in our team. We collect data from American Soccer Analysis. Our goal is seeking soccer players who are undervalued by performing statistical analysis, and eventually we would recommend a team of a group of players that could pull out the best performance and cost less. My contribution are performing statistical tests such as ANOVA and t-test to confirm our hypothesis and clustering players into different groups using K-means (page 11 - 15). Our final deliverable is the formation of new team with undervalued players.
Lending Club Loan Data Analysis
Technical skills: R, Exploraroty Data Analysis, Machine Learning
This is my project in Data Mining Graduate Course. I led to team to clean and deliver exploratory data analysis on Lending club loan data. The data set is used in Kaggle competition and is available to download here. My technical task is to perform clusterring analysis to segment customers into 4 groups based on their loan conditions and their credit scores. I also provided guidance and assisted other team members to use Random Forest and Discriminant algorithms to predict risk of late payment for current loans.
Operation Research Project
Stock Invesment Portfolio Optimization
Technical skills: R, Network Modelling, Linear Programming
This project is to apply optimization algorithms and techniques for stock investment. Our investment team will choose 40 stocks to invest and only consider buying or selling these stocks during the three-year-long investment period. The decision of buying and selling will based on the result of scenario tree with the predication on the economy of the United States and the network optimization to maximize the total revenue generated from the stock investment portfolio. Finally, there will be eight investment portfolios for different conditions and predictions. At the same time, to lower the risk, the stocks to be chosen are divides into two groups: better performance in good economy (positive beta compared to SP500 index-SPY) (Group 1) and better performance in bad economy (negative beta compared to SPY) (Group 2).
Time Series Analysis Project
Prediction of USD/CNY Exchange Rate
Technical skills: Eviews, ARIMA, Regression Modelling
This is my team project in FIN642: Business Conditions and Forecasting course, which covers comprehensive and state-of-the-art forecasting methods applied to the business world. We investigates the factors that affect exchange rate movements between USD and CNY and makes forecasting for future USD/CNY exchange rate.
The techniques we used includes various time series analyses such as ARIMA and Holt-Winters exponential smoothing and econometric modeling. The time series analysis suggests that the exchange rate change follows an autoregressive model using data from the most recent time period. The evidence from the regression model indicates that the change in exchange rate is positively correlated with the change in forward premium of the most recent time period. The root of mean squared error suggests that the time series model outperforms the regression model.
Consulting Project
Telematics Data Analysis
Technical skills: Tableau, SQL, Data Visualization
This is a consulting project that I did with a team of 5 graduate students for a client of LeBow Business Analytics Center. We investigated the best practices for the use of telematics data and analyzed data sets provied by the client to give useful insights for Supply Chain Department. My key responsibilities were conducting industry research, cleanning and aggregating data, visualizing and interpreting clustering analysis on idling time.
Forecasting and Optimizing Call Center Staffing
Technical skills: Time Series Analysis, Statistical Modeling, R
This is my Capstone Project in which I lead a team of 4 graduate students in the first 6 weeks and then team of 15 graduate students in remaining weeks to deliver a data modeling and statistical analysis for a partner of Drexel University. We were asked to build time series models to forecast call volume, call length and service level using in-house data from the last 2 years and then, calculate the number of agents required next year in the call center. I was in charge of communicating with clients, framing business problems and then providing guidance for my group. In term of technical tasks, I performed analysis to forecast service level, service time and shrink duration using multiple regression and moving average method. I was also in chare of researching queuing models and providing client with staff calculation models. Our team sucessfully presented to C-suite executives at client's headquarter and I, myself, got an A+ for my contribution and leadership.
