TABLE 1. Demonstration of the High Performance of Deep Learning Approach Over Machine Learning.
| Procedure | F1 score for Class | Accuracy | ||
|---|---|---|---|---|
| Class 0 | Class 1 | Class 2 | ||
| Random Forest | 0.95890 | 0.66315 | 0.68124 | 0.92689 |
| Naïve Bayes | 0.29003 | 0.02210 | 0.05677 | 0.18305 |
| KNN (K = 4) | 0.94294 | 0.33855 | 0.40343 | 0.88601 |
| KNN (K = 5) | 0.92958 | 0.33058 | 0.37620 | 0.86321 |
| KNN (K = 6) | 0.93416 | 0.33537 | 0.38710 | 0.87146 |
| CNN (1 layer) | 0.92760 | 0.40901 | 0.34286 | 0.85941 |
| CNN (2 layer) | 0.97529 | 0.71035 | 0.88293 | 0.95453 |
| Proposed | 0.98075 | 0.75703 | 0.93734 | 0.96502 |