<![CDATA[The objective of this study was to compare the efficacy of various machine learning models for classifying chemical substances using sensor array data from a wind tunnel facility. Six widely recognized machine learning algorithms were assessed: Random Forest, Gradient Boosting, Logistic Regression, Support Vector Machine (SVM), Decision Tree, and K-Nearest Neighbors (KNN). The dataset, consisting of 288 sensor array features, was preprocessed and utilized to evaluate the models based on accuracy, precision, recall, and F1 score through a 5-fold cross-validation method. The results indicated that ensemble methods, particularly Random Forest and Gradient Boosting, outperformed other models, achieving an accuracy and F1 score of over 99%. KNN also demonstrated high efficacy with similar performance metrics. In contrast, Logistic Regression showed modest results in comparison. The study's outcomes suggest that ensemble machine learning models are highly suitable for chemical detection tasks, potentially contributing to advancements in environmental monitoring and public safety. The findings also highlight the importance of quality data preprocessing in achieving optimal model performance. Future research directions include exploring hybrid models, deep learning techniques, and assessing model robustness against environmental variabilities. This research underscores the transformative potential of machine learning in chemical detection and paves the way for developing more sophisticated and reliable detection systems.]]>
Copyrights © 2024
