Service Project

Jason Wang

1-714-***-**** ********@*****.***

* / *

Skills Languages

• R(2yrs+), SQL, Python, Java(5yrs+), JavaScript, HTML5, CSS3, C++, Visual Basic.

Platforms

• TensorFlow, MySQL, MongoDB, AWS EC2, Apache Tomcat, Java servlets, JUnit, GitHub, etc.

Coursework

• Mathematical Statistics, Multivariate Statistics, Categorical Data Analysis, Statistical Consulting, Bayesian Statistics, etc.

• Data Structure and Object-Oriented Design in Java, Design and Analysis of Algorithms, Database System Principles, etc.

Education Microsoft Technology Associate 2017

• Software Development Fundamental Certification

M.S. in Statistics, California State University Fullerton 2015 - 2017

• Major GPA 3.97

B.S. in Computing Mathematics, City University of Hong Kong 2005 - 2009

• Full Tuition Scholarship and Dean’s List

Projects

Image Recognition Project – Python, TensorFlow

Aimed to develop a classifier which can identify a specific logo from a random set of Instagram photos.

• Researched common methods in image processing, such as histogram of oriented gradients(HOG) and convolutional neural network(CNN).

• Designed a binary classifier based on the methodology of CNN, and implemented the transfer learning idea by using TensorFlow.

• Utilized the regularization penalty and dropout technique during the training process, and improved the model performance to 90% in sensitivity and 87% in specificity for 200 test photos.

• Addressed the training efficiency issue via GPU-based accelerated computation on AWS EC2, and delivered the final classifier which can generate result within 0.3 seconds for a single image.

Customer Churn Prediction Project – Python, scikit-learn Aimed to construct a model which can predict the propensity to end service for the given customers.

• Implemented supervised learning methods, such as generalized linear regression (binomial case) and k-nearest neighbors algorithm.

• Resolved overfitting issue by applying ridge regression and random forest method.

• Selected the model by evaluating the classification performance via k-fold cross-validation.

• Performed lasso regression for feature selection, and analyzed top factors that influenced the user intention in service termination. Restaurant Search and Recommendation Project – Java, MySQL Aimed to develop a web application which can provide a list of restaurant recommendation based on user preference or location information. Jason Wang

1-714-***-**** ********@*****.***

2 / 2

Projects

(Continued)

• Researched common recommendation algorithms including user-based and item- based methods.

• Implemented a web service using Java servlet and REST API to fetch restaurant data from Yelp API.

• Utilized MySQL to store user information and restaurant information.

• Designed an item-based filtering and sorting algorithm to match contents, so that similar restaurants can be recommended to relevant users.

• Tested the web service and application with unit test(JUnit). Product Recommendation Project – Java, MongoDB

Aimed to provide the product recommendation for customs using Amazon’s dataset.

• Extracted and purified the metadata of product review from Amazon and prepared the dataset for testing.

• Implemented a MapReduce program in MongoDB to calculate similarity scores between users.

• Applied a collaborative filtering algorithm based on similarity scores to improve the recommendation efficiency.

• Tested the performance among 339,291 customs and 83,046 products. Work

History

COSCO Container Service Limited

Manager of Enterprise Management Division

2011 - 2015

• Collected operational data and converted to deliverable KPI dashboards. Presented the forecast reports in Group’s monthly progress meeting and participated the managerial decision making.

• Liaised with different departments to track the status of the company’s operational environment.

Convoy Financial Holdings Limited

Consultant

2009 - 2010

• Developed financial and investment strategies for private bank clients.

Contact this candidate