Li Du Email and Phone Number

• Develop statistical models and analyze time series of the investment and revenue o quantitatively measure the efficiency of various marketing activities across different brands• Scrape over 5000 customer reviews for the Movado watches on major E-commerce platforms and perform NLP analysis such as aspect-based sentiment modeling to understand the customers’ acceptance of the products• Query and analyze relational databases containing millions of customer and transaction information using SQL and Python to understand the customers’ behavior and provide suggestions for better segmentation and personalization.

• Explored various sources of data and performed QA/QC to the data obtained • Act as the lead developer of core model structures and machine learning algorithms for a data science product that predicts urban on-road traffic crash risks. The product has been purchased by a city department.• Explore machine learning algorithms to improve the model performance and to address the emerging needs from both existing and new clients.• Communicate and explain methodologies and findings in client meetings regarding technical aspects of projects.

• Provided insights to local regulators by using Python to analyze 15+ years of São Paulo (Brazil) metro area air quality monitoring data in support of clean vehicle and fuel policies in Brazil.• Researched and implemented algorithms and spatial surrogates in ArcGIS to dis-aggregate country-level diesel vehicle emission data into grid cells for subsequent global air quality modeling; results were published in world-leading science journal (Nature) in response to Volkswagen’s emissions scandal (Dieselgate).• Spearheaded a $50K+ project that modeled greenhouse gas emissions and costs associated with seven phase-in strategies for climate-friendly refrigerants in Chinese light-duty vehicle fleet.

• lead a 5-year air quality monitoring program and perform QA/QC to air quality data collected from the monitoring station.• optimize a non-parametric wind regression model and perform analysis to identify air pollutant emission sources• compile and process 15+ years of air quality monitoring data (PM2.5 and gaseous air toxics) from 5 stations in the St. Louis metropolitan area for the quantitative trend analysis of air quality• test and optimize the positive matrix factorization (PMF) model; quantitatively identify and model the sources and their contributions of PM2.5 in the St. Louis area• Characterized measurement errors within air monitoring networks across the United States by applying advanced statistical methods (Bootstrapping, Jackknife etc.) in R to datasets consisting of millions of air quality data records.• apply GIS-based approaches to estimate air pollution exposure for patients enrolled in a research program that aimed at linking air pollution to the inception of asthma after Respiratory syncytial virus (RSV) Bronchiolitis in early life (in collaboration with Washington University, School of Medicine)

• Conducting field campaigns in rural areas of China to study the indoor air quality and cook stove emissions;• Conducting field sampling campaigns in metal-processing industries and power plants for the investigation of particulate and mercury emissions.

• Working as a project co-manager for the "Evaluation of the air quality Impact by the Installation of a New Coking Facility at Taiyuan Iron & Steel Taiyuan Works". Responsible for the planning of air sampling campaign, on-site sampling operations coordination as well as data analysis;• Performing optimization to the desulfurization facility at Shijiazhuang Iron & Steel Works.

Testing indoor air pollutants levels at multiple residential and public locations and working as a consultant for the improvement of indoor air quality.