Data Analyst Assistant

Jason (Weinan) Zhi

Worcester, MA • 508-***-**** • *********@*****.*** • GitHub: https://github.com/WeinanZhi• LinkedIn: https://www.linkedin.com/in/weinan-zhi-7568b865

Data Analyst who enjoys driving business decisions. Experience designing queries, implementing non-linear dimension reduction methods and deep learning. Strengths in effective leadership, communication, driving insights and improving results.

AREAS OF EXPERTISE

Deriving actionable insights and translating complex results or algorithms into simple conclusions that allows for actionable data-driven decisions and solutions.

Data extraction, profiling, cleansing, deduplication, standardization, conversion, transformation, data mining, and reporting.

Requirements and data gathering, integration and enrichment.

Business intelligence methods focused on developing dashboards.

Predictive modeling Bayesian statistics

TECHNICAL PROFICIENCIES

Python, SQL, MySQL, Oracle, R, Linux, XML, Tableau, Java, GitHub

Professional Experience

Mapfre Insurance Company, Webster, MA

GQP Contractor (02/2020 – 05/2020)

Improved the productivity and efficiency of communication by designing KPIs and performing time series analysis.

Delivered a new GUI, derived by generating employee clusters, that displayed detailed personal information.

Adopted ARIMA models as baseline models for a time series analysis on Microsoft Office data.

Shanghai Yutai Investment Co, LTD, Pudong, Shanghai

Assistant to the CEO (08/2016 – 04/2018)

Designed and developed the company’s website

Highland Park Historical Society, Highland Park, Chicago, Illinois

Intern (07/2014 – 08/2014)

Digitized archives and optimized current website.

Data Projects

Pneumonia Detection on Chest X-ray - GitHub: https://github.com/WeinanZhi/Pneumonia-Detection-on-X-ray

Drastically reduced the dimension of chest x-ray images by ISOMAP and Laplacian Eigenmaps algorithm to solve a classification problem on Pneumonia. Multiple layer perceptron and logistic regression models were applied as evaluation. Both of the two classifiers received a 0.86 test F1 score.

GoogleMap Image Geolocation – GitHub: https://github.com/snugfox/aerial-image-classification

Implemented transfer learning of ResNet50/VGG16 pretrained layers to geolocate the created balance dataset of 5000 satellite images generated by the GoogleMaps API. Reached 45.6%/42.8% classification accuracy on country level.

Plant Seed Prediction - GitHub: https://github.com/WeinanZhi/-CS539-PlantSeedPrediction

Implemented transfer learning of ResNet18/DenseNet121 pretrained layers to classify created balanced dataset of segmented unsegmented plant seed images. Reached 87.6%/92.4classificatonccuracy on average.

Indexing Improvement on SimpleDB – GitHub: https://github.com/WeinanZhi/SimpleDB_3.00

Optimized the range query by implementing a R tree algorithm as well as optimized the hash index in SimpleDB by a linear hash index. Also, designed 10 query plans to evaluate improvement performance.

Education

Master of Science in Data Science, Worcester Polytechnic Institute, Worcester, MA, 3.63

Bachelor of Management in Information Management & Information System, Renmin University of China (RUC), Beijing

Contact this candidate