Thesis

3 65 MACHINE LEARNING SUPPORTING SUBSTITUTIONS IN SOCCER Table 3. Machine learning metrics. Variable: Distance Covered Threshold Algorithm Accuracy AUC Precision Recall F1-score 100% Random Forest 0.90 0.94 Underperforming 0.97 0.92 0.94 Performing 0.70 0.84 0.76 Decision Tree 0.88 0.86 Underperforming 0.96 0.90 0.93 Performing 0.64 0.82 0.72 Naïve Bayes 0.57 0.58 Underperforming 0.83 0.59 0.69 Performing 0.21 0.48 0.29 95% Random Forest 0.75 0.79 Underperforming 0.71 0.67 0.69 Performing 0.78 0.81 0.79 Decision Tree 0.77 0.82 Underperforming 0.73 0.72 0.72 Performing 0.80 0.81 0.81 Naïve Bayes 0.73 0.75 Underperforming 0.68 0.67 0.67 Performing 0.77 0.78 0.77 90% Random Forest 0.93 0.95 Underperforming 0.55 0.87 0.67 Performing 0.99 0.94 0.96 Decision Tree 0.92 0.88 Underperforming 0.51 0.85 0.63 Performing 0.99 0.93 0.96 Naïve Bayes 0.74 0.74 Underperforming 0.15 0.52 0.24 Performing 0.95 0.92 0.84 Variable: Distance In Speed Category Model Threshold Algorithm Accuracy AUC Precision Recall F1-score 100% Random Forest 0.89 0.96 Underperforming 0.85 0.87 0.86 Performing 0.91 0.90 0.91 Decision Tree 0.74 0.81 Underperforming 0.65 0.73 0.68 Performing 0.81 0.75 0.78 Naïve Bayes 0.70 0.78 Underperforming 0.59 0.77 0.67 Performing 0.82 0.65 0.72 95% Random Forest 0.96 0.98 Underperforming 0.68 0.91 0.78 Performing 0.99 0.96 0.98 Decision Tree 0.94 0.92 Underperforming 0.59 0.89 0.71 Performing 0.99 0.95 0.97 Naïve Bayes 0.97 0.83 Performing 0.23 0.72 0.35 Performing 0.97 0.80 0.87 90% Random Forest 1.00 0.99 Underperforming 0.61 0.94 0.74 Performing 1.00 1.00 1.00 Decision Tree 0.99 0.97 Underperforming 0.41 0.94 0.57 Performing 1.00 0.99 1.00 Naïve Bayes 0.88 0.89 Underperforming 0.03 0.74 0.06 Performing 1.00 0.88 0.93 Variable: Energy Expenditure In Power Category Threshold Algorithm Accuracy AUC Precision Recall F1-score 100% Random Forest 0.89 0.96 Underperforming 0.88 0.89 0.89 Performing 0.89 0.89 0.89

RkJQdWJsaXNoZXIy MjY0ODMw