لخّصلي

This section presents the outcomes of employing various machine learning models for detecting fire extinguisher devices using acoustic waves.Other models, including Naive Bayes, KNN, Logistic Regression, Decision Trees, and SVM, also demonstrate competitive performance, albeit with slight variations in effectiveness.Figure 5 and Table 1 depict the accuracy metrics for each model, namely Logistic Regression, Random Forest, Support Vector Machine (SVM), K-Nearest Neighbor (KNN), and Naive Bayes.The results show variations among the models, with Random Forest achieving the highest accuracy of 98.33%, followed closely by Logistic Regression with 97.33%, SVM with 97.33%, Naive Bayes with 96.33%, and KNN with 95.83%.The other models, including Logistic Regression, Decision Trees (DT), KNN, and Naive Bayes, showcased precision scores ranging from 92.53% to 97.41%.Conversely, Logistic Regression and Support Vector Machine displayed lower recall rates at 90.33%.Random Forest attained the highest F1 score of 98.34%, followed by Logistic Regression at 92.34%.Other models, such as Decision Trees, SVM, KNN, and Naive Bayes, exhibited F1 scores ranging from 91.76% to 95.34%.

This section presents the outcomes of employing various machine learning models for detecting fire extinguisher devices using acoustic waves. It encompasses the preliminary findings of the research and offers recommendations based on the results obtained. The evaluation of the effectiveness of these models is conducted through various metrics including accuracy, precision, recall, and F1 score.

Figure 5 and Table 1 depict the accuracy metrics for each model, namely Logistic Regression, Random Forest, Support Vector Machine (SVM), K-Nearest Neighbor (KNN), and Naive Bayes. The results show variations among the models, with Random Forest achieving the highest accuracy of 98.33%, followed closely by Logistic Regression with 97.33%, SVM with 97.33%, Naive Bayes with 96.33%, and KNN with 95.83%.

Precision metrics, illustrated in Figure 6 and Table 2, highlight the capability of each model to correctly identify positive cases. Support Vector Machine exhibited the lowest precision at 91.41%, while Random Forest demonstrated the highest precision at 98.41%. The other models, including Logistic Regression, Decision Trees (DT), KNN, and Naive Bayes, showcased precision scores ranging from 92.53% to 97.41%.

Figure 7 and Table 3 present the recall metrics, indicating the ability of the models to identify all relevant instances. Random Forest and Decision Trees exhibited the highest recall scores at 98.33% and 97.5%, respectively. Conversely, Logistic Regression and Support Vector Machine displayed lower recall rates at 90.33%. KNN and Naive Bayes showed recall rates of 93.83% and 94.33%, respectively.

The F1 scores, depicted in Figure 8 and Table 4, provide a balance between precision and recall. Random Forest attained the highest F1 score of 98.34%, followed by Logistic Regression at 92.34%. Other models, such as Decision Trees, SVM, KNN, and Naive Bayes, exhibited F1 scores ranging from 91.76% to 95.34%.

Table 5 summarizes the average performance of the machine learning models across all metrics. Random Forest emerges as the top-performing model, consistently achieving high scores in accuracy, precision, recall, and F1 score. Other models, including Naive Bayes, KNN, Logistic Regression, Decision Trees, and SVM, also demonstrate competitive performance, albeit with slight variations in effectiveness.