Menglu Huang Email & Phone Number

Boulder, Colorado, Us

• Spearheaded large-scale data science projects involving in-depth data validation, statistical analysis, and predictive analytics that drive revenue improvement and operational efficiency• Specialized in people analytics segmentation, risk analysis, propensity enrollment model and energy usage load profiling• Experienced in handling large-scale data engineering, high-dimensional feature selections, hyperparameter tuning•Leveraged advanced natural language processing (NLP) techniques and algorithms for text classification, information extraction, and sentiment analysis on survey data and demographics data.• Built time series analysis pipeline that incorporated hierarchy of new business projects across multiple levels of service territory; monitored the entire DS life cycle of modeling through weekly sprints and code review

Princeton, Nj , Us

Developed and documented programs that clean, transform, and combine datasets with over 1 million observations;Provided analytical support and technical insights for program evaluations by analyzing datasets;Contributed to 6-month billing and payment analysis and identified key performance indicator for energy programs.

Pittsburgh, Pennsylvania, Us

• Performed in-depth data validation spanning over 30 million student learning data• Analyzed large-scale students’ empirical performance data using SQL and Hadoop Hive• Performed data visualization on student performance using ggplot2 in R• Modeled actual learning outcomes across 458 learning activities using hierarchical logistic regression and identifiedmismatches between algorithm output and empirical learning results• Provided statistical consulting for clients at Open Learning Initiative Education Platform on their current algorithms for learning productivity estimation

Pittsburgh, Pa, Us

San Francisco, Ca, Us

• Conducted research on associated restaurants’ portfolio and spent 80+ hours on exploratory data analysis• Constructed an estimation model based on the company strategy, historical data, and industry trends to predict wait time • Used random forest and gradient boosting algorithm to train data and successfully increased accuracy of the current model.

Pittsburgh, Pa, Us

The Pennsylvania Alternate System of Assessment and Students with Disabilities and Challenging Behavior: What Teacher Strategies Support Better Student PerformanceCompiled cross-sectional data to examine relationship between teaching strategy and student behaviorsAnalyzed data of students with disabilities and challenging behaviors through codebookTheorized suitable teaching methods according to research results, which were integrated into assessment

Master'S Degree, Statistics

Bachelor'S Degree, Economics

High School Diploma

High School Diploma