Liquid Classification based on MMW experiments and Multiclass SVM
Résumé
An innovative approach to liquid classification by combining Support Vector Machine (SVM) with cross-validation is proposed. Using a database extended by patch extraction, the model achieves 100% performance in differentiating various liquids in different configurations. The consistency of the results, highlights the robustness of the model, no matter the specific composition of the learning base. This method offers important implications for the safety of public spaces by providing an advanced solution for the accurate detection of liquids. Moreover this promising approach could be extended to more complex scenarios in the future, reinforcing its applicability.
Mots clés
Support vector machines
Training
Liquids
Accuracy
Databases
Machine learning
Robustness
Support Vector Machine
Multi-class Support Vector Machine
Complex Scenarios
Patch Extraction
Combination Of Support Vector Machine
Machine Learning
Pedestrian
Cross-validation Method
Energy Drinks
Liquid Type
Grey Wolf Optimizer