Figure 6
The figure contains two rows and two columns of plots labeled (a) through (d), comparing hyperparameter importance and optimization history for two datasets: “M C S I” and “M S L D v 2”. Panel (a), “Param importance (M C S I)” (left, top row): A horizontal bar chart titled “Hyperparameter Importances”. The horizontal axis is labeled “Hyperparameter Importance” and runs from 0.0 to about 0.6 with an interval of 0.1. The vertical axis is labeled “Hyperparameter” and lists four parameters: “kernel”, “C”, “degree”, and “gamma”. The bars show kernel: 0.60 (longest bar), C: 0.20, degree: 0.11, and gamma: 0.08. Panel (b), “Opt history (M C S I)” (right, top row): A scatter-and-line plot titled “Optimization History Plot”. The horizontal axis is labeled “Trial” and runs from 0 to 14 with an interval of 2. The vertical axis is labeled “Objective Value” and ranges from 0.7 to 1 with an interval of 0.05. Circular data points labeled “Objective Value” are plotted across trials, and a line labeled “Best Value” tracks the best value so far. The best-value line jumps from about 0.76 at trial 0 to about 0.98 at trial 1, then remains near 0.98 with a small increase to about 0.99 around trial 11, staying flat through trial 14. Most blue points lie between about 0.97 and 0.99, with two noticeable low outliers near trial 9 (around 0.70) and trial 10 (around 0.78). Panel (c), “Param importance (M S L D v 2)” (left, bottom row): A horizontal bar chart titled “Hyperparameter Importances”. The horizontal axis is labeled “Hyperparameter Importance” and runs from 0.0 to about 0.7 with an interval of 0.1. The vertical axis is labeled “Hyperparameter” with the same four parameters: “kernel”, “C”, “degree”, and “gamma”. The bars show kernel: 0.73 (longest bar), C: 0.15, degree: 0.06, and gamma: 0.06. Panel (d), “Opt history (M S L D v 2)” (right, bottom row): A scatter-and-line plot titled “Optimization History Plot”. The horizontal axis is labeled “Trial” and runs from 0 to 14 with an interval of 2. The vertical axis is labeled “Objective Value” and ranges from near 0.2 to 1 with an interval of 0.2. Circular data points labeled “Objective Value” appear across trials, and a “Best Value” line shows the best value so far. The best-value line starts near 0.93 at trial 0, rises to about 0.99 at trial 1, and stays flat close to 0.99 through trial 14. Most objective-value points lie between about 0.85 and 1.0, with two very low outliers around trials 2 and 4 near 0.08 and a lower point around trial 14 near 0.83. Note: All the numerical data values are approximated.Hyperparameter tuning results for both datasets showing parameter importance and optimization progress