Sign in

Project Manager Data Analyst

Berkeley, CA
March 07, 2018

Contact this candidate


Yan Zeng

**** ********** ******, *** ***, Berkeley, CA, 94704 651-***-****


University of California, Berkeley, CA Expected May 2018 Master of Engineering in Industrial Engineering & Operations Research GPA: 3.97/4.00 Relevant Coursework: Data Analysis, Risk Modeling, Simulation, Optimization University of Wisconsin, Madison, WI May 2017

Bachelor of Science in Industrial Engineering with Highest Distinction GPA: 3.85/4.00 Certificates: Secondary major in Mathematics, Computer Science Certificate, Six Sigma Green Belt Certificate SKILLS

§ Technical Skills: Clustering, Accounting and Finance, Stochastic Modeling, Lean Six Sigma, Project Management

§ Machine Learning Skills: Lasso, Ridge, Logistic Regression, Random Forest, KNN, K-Means, SVM, Hierarchical Clustering, PCA

§ Programming Languages/Computer Skills: Java, Python, R, SQL, MATLAB, Arena, Visio, Maple, Microsoft Basic

§ Problem Solving Tools: DMAIC methodologies, PDCA analysis, 5W1H, Fishbone, 5Why, Processing Map, SIPOC WORK EXPERIENCE

Data Analyst: Design a Skin Disorder Prediction Model, Capstone Project UC San Francisco Liao’s Lab, Berkeley, CA May 2017 – May 2018

§ Detected outliers, imputed missing values, and checked multicollinearity to conduct feature engineering on 86000 records

§ Applied supervised learning models to predict top 10 factors that led to Psoriatic Arthritis (a skin disorder) Project Manager in Patients’ Waiting Time Reduction Consulting UW Health Fitchburg Clinic, Madison, WI Jan 2017 – May 2017

§ Led an engineering team of three to decrease the average waiting time of the Biopsy Test result-receiving process from over 6 weeks to within 4 weeks using DMAIC and PDCA methodologies

§ Utilized excel pivot table to analyze 6 months of patient records and suggested solutions that managed the risk of delays

§ Assigned tasks to each team member and ensured deliverables are completed before the deadlines

§ Proposed biopsy test standard protocol for the clinic staff members to minimize the lead times in the process

§ Presented solutions to over 20 clinic physicians in hospital and achieved satisfaction from higher management groups Project Manager in Process Improvement Consulting

UW University Health Services, Madison, WI Sept 2016 – Dec 2016

§ Identified 10 sources of inefficiency and errors in the Tuberculosis skin testing process and formulated 5 implementation plans and recommendations that are feasible for on-site team to complete

§ Coordinated with 6 departments to arrange weekly meetings for process update and for answering questions

§ Interviewed over 100 end users and senior management to collect feasibility assessment and other critical data

§ Led a team of four to design the final solution that decreased percentage of patients who went to the wrong location in the process from 2.94% to 2% focusing on the error prevention of the system PROJECTS EXPERIENCE

Data Analyst: Design a Hiring Tool for Human Resources Aug 2017 – Dec 2017

§ Verified feature correlations and performed PCA and data normalization to clean and transform over 2000 survey records

§ Constructed 6 supervised and 4 unsupervised models to understand each individual’s personality and behavior

§ Tuned and validated models and picked the discrete K-Means model based on interpretability and accuracy

§ Used Django framework deployed on Heroku to build a web-based user interface to auto-generate the survey outcome Data Analyst: Identify Bottlenecks on Low Website Conversion Rate May 2017 – Aug 2017

§ Performed data preprocessing to extract 70 features and imputed missing data with feature engineering techniques

§ Identified key factors that bottleneck the conversion rate of the website through building the hypothesis experiments

§ Implemented Logistic Regression and RF models to predict top 10 influential factors that will affect users’ signup behavior

§ Models yielded high performance with AUC score of 0.97 and 96.67% accuracy on test data Data Analyst: Yelp Data Challenge Analysis May 2017 – Aug 2017

§ Used tokenization with stemming and lemmatization to convert user data to vector space for NLP study

§ Built a Logistic Regression model based on user tips and reviews feature to predict successfulness of a business entity

§ Performed unsupervised learning to cluster users and analyzed common user preferences by inspecting the centroid

§ Applied collaborative filtering to build a restaurant recommender with accuracy of 85%

Contact this candidate