• Developed 120+ Test Scenarios and Test Cases using SQL queries to validate data accuracy (Sales, Member Points, Promotions) to ensure data is correct according to business logic requirement.
• Collaborated with clients to execute User Acceptance Testing (UAT) for CRM, Loyalty, and Sales Retail, achieving 99% consistency between data source and new data structures.
• Investigated data inconsistencies (e.g., calculation logic, source errors) and proposed solutions to obtain client sign-off, delivering accurate and ready-to-use data for BI reporting.
ประวัติการฝึกอบรม
สิงหาคม 2566 ถึง สิงหาคม 2566
Government Big Data Institute - GBDi
Hands-on Data Science and Machine learning
กันยายน 2566 ถึง กันยายน 2566
Government Big Data Institute - GBDi
Intermediate Data Engineering
ความสามารถ
ความสามารถทางภาษา
พูด (ดีมาก) อ่าน (ดีมาก) เขียน (ดีมาก)
พูด (พอใช้) อ่าน (ดี) เขียน (พอใช้)
ไทย คำ/นาที อังกฤษ คำ/นาที
โครงการ / ผลงาน / เกียรติประวัติ / บุคคลอ้างอิง
1. Online Shopper Intention Prediction | Python, Scikit-learn
• Developed a Classification Model to forecast customer purchasing intent, achieving 90% Accuracy validated by 10-Fold Cross-Validation to verify model stability.
• Optimized model performance to attain 74% Precision and 65% F1-Score by analyzing key behavioral metrics (e.g., Page Values, Bounce Rates), effectively identifying high-potential buyers.
2. Heart Disease Classification | Python, K-NN, EDA
• Built a K-Nearest Neighbors (K-NN) classifier using Scikit-learn to predict heart disease risk, achieving a maximum Accuracy of 86.84% through Hyperparameter Tuning to optimize the K-value.
• Conducted Exploratory Data Analysis (EDA) using Matplotlib to visualize data distributions and correlations, facilitating feature selection and optimal 'K-value' determination.
3. Social Media Engagement Analysis
• Conducted comparative engagement analysis, finding that morning posts (6 AM–12 PM) outperform night posts by 73.25%, leading to data-driven recommendations for optimal posting schedules.
• Analyzed keyword performance and discovered that humorous keywords (e.g., “555”) drive 42.45% higher engagement, demonstrating the impact of emotional cues on content effectiveness.
4. Wat Phon Mani Website Development
• Designed Relational Database architecture and ER Diagrams based on requirement gathering, resulting in a system prototype that supported dynamic content management and satisfied user needs.