ZHILIANG WANG

Technical Consultant @ 凯捷中国

About

I am a motivated M.S. student in Electrical Engineering at Columbia University, passionate about using technology to improve human life. I have strong foundations in data structures, algorithms, machine learning, and distributed computing, and actively pursue opportunities in software development, data analysis, and AI solutions. I have practical experience in Python, C++, AWS, and ML frameworks: building credit scoring and insurance prediction models, optimizing data management systems, automating workflows, and researching black-box adversarial attacks. Beyond technical skills, I led a volunteer teaching project in Guizhou, introducing children to computer knowledge and e-learning resources, fostering social impact alongside technical innovation. I am eager to apply my analytical mindset, software expertise, and collaborative spirit to tackle real-world problems in data and software domains.

Country

United States

City

New York

Industry

Information Technology & Services

Skill

Data Reduction, Predictive Modeling, Semiconductor Engineering, Statistical Modeling, Research Implementation, Database Scripting, Distributed Caching, Word Embeddings, Machine Learning, Intrusion Detection, Database Optimization, Real-time Web, AWS CodeBuild, AWS Step Functions, Statistical Data Analysis, Data Cleaning, Algorithm Design, Data Visualization, Data Transformation Services (DTS), Data Monitoring

Experience

Technical Consultant

凯捷中国

2024-7 - 2024-10 · 4 mos

中国广东省深圳

Devised a Python-based customer credit scoring system on the German Credit dataset, applying WOE feature engineering and random forest, and fine-tuning ensemble ML models; improved model accuracy by 12% and AUC by 0.08, producing more reliable scorecards for credit risk assessment. Created a Python-based insurance premium prediction model on an insurance dataset, performing data cleaning, statistical analysis, and applying ML algorithms (KNN, etc.), revamped R² score from 0.72 to 0.83, enhancing premium estimation accuracy. Enhanced a Python-based data management system for a Meituan takeout site by resolving rider labeling errors and leveraging data visualization to analyze delivery performance, providing actionable improvement plans.