10
folds and 1
repetitions.The performance is measured in three different ways.
algo | model | succ | par10 | mcp |
---|---|---|---|---|
baseline | vbs | 1.000 | 1.000 | 0.000 |
baseline | singleBest | 1.000 | 0.959 | 0.041 |
baseline | singleBestByPar | 1.000 | 0.065 | 0.935 |
baseline | singleBestBySuccesses | 1.000 | 0.065 | 0.935 |
classif | rpart | 0.999 | 1.001 | 0.012 |
classif | randomForest | 0.999 | 1.002 | 0.010 |
classif | ksvm | 1.000 | 0.959 | 0.041 |
cluster | XMeans | 1.000 | 0.982 | 0.018 |
regr | lm | 0.999 | 0.987 | 0.028 |
regr | rpart | 1.000 | 0.980 | 0.020 |
regr | randomForest | 1.000 | 0.991 | 0.011 |
The following default feature steps were used for model building:
ALL
Number of presolved instances: 0
The cost for using the feature steps (adapted for presolving) is: 0
or on average: NA
The feature steps correspond to the following 55
/ 55
instance features:
CAPACITYOFKNAPSACK, DIMENSION, KNAPSACKDATATYPE, NUMBEROFITEMS, ITEMSPERCITY,
RENTINGRATIO, MAXSPEED, MINSPEED, angle_min, angle_mean,
angle_median, angle_max, angle_sd, centroid_centroid_x, centroid_centroid_y,
centroid_min_distance_to_centroid, centroid_mean_distance_to_centroid, centroid_max_distance_to_centroid, cluster_01pct_number_of_clusters, cluster_01pct_mean_distance_to_centroid,
cluster_05pct_number_of_clusters, cluster_05pct_mean_distance_to_centroid, cluster_10pct_number_of_clusters, cluster_10pct_mean_distance_to_centroid, chull_area,
chull_points_on_hull, distance_min, distance_mean, distance_median, distance_max,
distance_sd, distance_distances_shorter_mean_distance, distance_distinct_distances, distance_mode_frequency, distance_mode_quantity,
distance_mode_mean, distance_mean_tour_length, modes_number, mst_depth_min, mst_depth_mean,
mst_depth_median, mst_depth_max, mst_depth_sd, mst_dists_min, mst_dists_mean,
mst_dists_median, mst_dists_max, mst_dists_sd, mst_dists_sum, nnds_min,
nnds_mean, nnds_median, nnds_max, nnds_sd, nnds_varcoef
21
to 1
, resulting in a PAR10 score of 0.065
for the reduced model. Analogously, the model that was generated based on 1
of the originally 52
features resulted in a PAR10 score of 0.959
.
Below, you can find the list of the selected features and solvers:
Selected Features:
KNAPSACKDATATYPE
Selected Solvers:
SH