diss_by_feature_fr.png
Table 28: SVM and RandomForest accuracy by feature in French

The image displays a data table titled Table 28, which compares the accuracy of two machine learning models, SVM (Support Vector Machine) and RandomForest, across various features for the French language. The table is organized into four columns.

The column headers are:
1. Features
2. fr.svm.acc (French SVM Accuracy)
3. fr.rf.acc (French Random Forest Accuracy)
4. avg.fr.acc (Average French Accuracy)

Below the headers, the table lists specific features and their corresponding numerical accuracy scores. The rows are sorted in descending order based on the average accuracy column.

Here is the data row by row:

Feature: A3P0 | fr.svm.acc: 76.824 | fr.rf.acc: 68.317 | avg.fr.acc: 72.571
Feature: Duration | fr.svm.acc: 65.787 | fr.rf.acc: 79.208 | avg.fr.acc: 72.497
Feature: Amp_F3 | fr.svm.acc: 76.788 | fr.rf.acc: 67.657 | avg.fr.acc: 72.222
Feature: A1P0_HighPeak | fr.svm.acc: 75.798 | fr.rf.acc: 67.070 | avg.fr.acc: 71.434
Feature: A1P0_Compensated | fr.svm.acc: 75.138 | fr.rf.acc: 66.300 | avg.fr.acc: 70.719
Feature: Amp_F1 | fr.svm.acc: 74.697 | fr.rf.acc: 64.173 | avg.fr.acc: 69.435
Feature: LowZeroDiffP0 | fr.svm.acc: 72.094 | fr.rf.acc: 60.946 | avg.fr.acc: 66.520
Feature: Width_F1 | fr.svm.acc: 69.380 | fr.rf.acc: 60.946 | avg.fr.acc: 65.163
Feature: Width_F3 | fr.svm.acc: 68.720 | fr.rf.acc: 58.306 | avg.fr.acc: 63.513
Feature: Amp_F2 | fr.svm.acc: 66.630 | fr.rf.acc: 59.406 | avg.fr.acc: 63.018
Feature: Freq_F3 | fr.svm.acc: 66.813 | fr.rf.acc: 55.996 | avg.fr.acc: 61.404
Feature: SpectralCOG | fr.svm.acc: 64.356 | fr.rf.acc: 57.462 | avg.fr.acc: 60.909
Feature: H1MinusH2 | fr.svm.acc: 63.806 | fr.rf.acc: 57.242 | avg.fr.acc: 60.524
Feature: P0Prominence | fr.svm.acc: 64.210 | fr.rf.acc: 56.766 | avg.fr.acc: 60.488
Feature: Width_F2 | fr.svm.acc: 64.173 | fr.rf.acc: 56.142 | avg.fr.acc: 60.158
Feature: Amp_P2 | fr.svm.acc: 63.623 | fr.rf.acc: 53.905 | avg.fr.acc: 58.764
Feature: Freq_F1 | fr.svm.acc: 56.032 | fr.rf.acc: 59.369 | avg.fr.acc: 57.701
Feature: LowZeroDiffA1 | fr.svm.acc: 61.643 | fr.rf.acc: 53.722 | avg.fr.acc: 57.682
Feature: Amp_P1 | fr.svm.acc: 59.699 | fr.rf.acc: 54.969 | avg.fr.acc: 57.334
Feature: Ratio_F2F3 | fr.svm.acc: 60.249 | fr.rf.acc: 54.089 | avg.fr.acc: 57.169
Feature: Freq_F2 | fr.svm.acc: 59.773 | fr.rf.acc: 54.309 | avg.fr.acc: 57.041
Feature: A1P1 | fr.svm.acc: 57.902 | fr.rf.acc: 53.319 | avg.fr.acc: 55.611
Feature: P1Prominence | fr.svm.acc: 55.996 | fr.rf.acc: 52.219 | avg.fr.acc: 54.107
Feature: A1P1_Compensated | fr.svm.acc: 58.379 | fr.rf.acc: 49.065 | avg.fr.acc: 53.722
Feature: MidZeroDiffP1 | fr.svm.acc: 56.106 | fr.rf.acc: 51.192 | avg.fr.acc: 53.649
Feature: MidZeroDiffA1 | fr.svm.acc: 54.785 | fr.rf.acc: 50.898 | avg.fr.acc: 52.842
Feature: A1P2 | fr.svm
This description was generated automatically. Please feel free to ask questions if you have further questions about the nature of the image or its meaning within the presentation.