Welding Quality Classifier using Optimized KNN Algorithm

Advanced welding quality classification system using optimized K-Nearest Neighbors algorithm. Predicts welding specimen performance based on carbon content and welding current, comparing traditional KNN with distance-limited variant for improved efficiency.

🔹 Algorithm Innovation: Traditional KNN vs Distance-Limited KNN comparison 🔹 Welding Expertise: Carbon content and current parameter optimization 🔹 Dual Visualization: Side-by-side algorithm performance comparison 🔹 Web Scraping: Multi-source data collection from specialized databases 🔹 Quality Control: Predicts optimal vs deficient welding outcomes

Distance-limited KNN reduces computational overhead

Visual boundary analysis for classification decision regions

Carbon content normalization for improved accuracy

Real-time classification with parameter adjustment capability

Industry Application: Enables welding engineers to optimize parameters before production, reducing defective welds and improving construction quality control.

Data Engineering: Web scraping pipeline from specialized welding databases, data cleaning and normalization for multi-source integration.