Peak-Based Machine Learning for Plastic Type Classification in Time-of-Flight Secondary Ion Mass Spectrometry

• DOI: 10.1021/jasms.4c00325
• Publication Date: 2024-11-08T17:39:12Z
• Authors: Jin Gyeong Son; Hyun Kyong Shon; Ji-Eun Kim; In−Ho Lee; Tae Geol Lee
• Keywords: Time-of-Flight; Boosting

Time-of-flight secondary ion mass spectrometry (ToF-SIMS) measurement data and machine learning were used in this work to classify six different types of plastics. In order take into account the characteristics data, local maxima first examined a preprocessing step. Several methods then implemented create model that could successfully To visualize distribution, we applied dimensionality reduction method, namely, principal component analysis. Finally, distinguish between plastics, conducted an ensemble analysis using four tree-based algorithms: decision tree, random forest, gradient boosting, LIGHTGBM. This approach can identify feature importance plastic samples allow inference chemical properties each type. way, ToF-SIMS be utilized plastics enhance explainability.