Jiamei Zeng

Master's

Data Analytics

Operations Research

Microsoft

Edge Mobile Data Analysis Intern

- Descriptive Analysis - LLM based Feedback Processing: Developed prompts and scripts to automate negative feedback classification via LLM API - User Behavior Analysis: Extracted multi-dimensional information (page flows, actions, tasks) from action data and developed a - Macro Data Analysis: Created a penetration index based on country-specific data like population and GDP, providing deeper - Weekly Data Reporting: Track key metrics (DAU, Retention, Install), and provided insights during weekly meetings - Predictive Analysis - User Churn Analysis: Conceptualized an analysis plan, labeled users as churned or retained by activity, utilized Tf-idf and logistic - Growth Forecasting: Modeled daily active user trends leveraging Prophet and other tools, drawing on year-over-year data and retention rates - Prescriptive Analysis - Advertising Budget Plan: Devised a budget strategy focusing on five-year profit maximization, based on historical advertising ROI - Causal Inference Analysis - Experiment Method Research: Proposed that cyclic user behavior led to fluctuations in Day 1 retention rates, affecting result