10
folds and 1
repetitions.The performance is measured in three different ways.
algo | model | succ | par10 | mcp |
---|---|---|---|---|
baseline | vbs | 0.875 | 6344.251 | 0.000 |
baseline | singleBest | 0.858 | 7201.556 | 79.143 |
baseline | singleBestByPar | 0.858 | 7201.556 | 79.143 |
baseline | singleBestBySuccesses | 0.858 | 7201.556 | 79.143 |
classif | rpart | 0.872 | 6488.162 | 10.512 |
classif | randomForest | 0.868 | 6677.876 | 22.360 |
classif | ksvm | 0.867 | 6727.606 | 27.624 |
cluster | XMeans | 0.862 | 7005.380 | 60.833 |
regr | lm | 0.869 | 6633.481 | 22.431 |
regr | rpart | 0.863 | 6952.248 | 52.167 |
regr | randomForest | 0.871 | 6540.345 | 18.228 |
The following default feature steps were used for model building:
all_feats
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 86
/ 86
instance features:
stats_varcount, stats_var_bool, stats_var_discrete, stats_var_bound, stats_var_sparsebound,
stats_dom_0, stats_dom_25, stats_dom_50, stats_dom_75, stats_dom_100,
stats_dom_mean, stats_dom_not2_2_ratio, stats_discrete_bool_ratio, stats_branchingvars, stats_auxvars,
stats_auxvar_branching_ratio, stats_conscount, stats_arity_0, stats_arity_25, stats_arity_50,
stats_arity_75, stats_arity_100, stats_arity_mean, stats_arity_mean_normalised, stats_cts_per_var_mean,
stats_cts_per_var_mean_normalised, stats_alldiff_count, stats_alldiff_proportion, stats_sums_count, stats_sums_proportion,
stats_or_atleastk_count, stats_or_atleastk_proportion, stats_ternary_count, stats_ternary_proportion, stats_binary_count,
stats_binary_proportion, stats_reify_count, stats_reify_proportion, stats_table_count, stats_table_proportion,
stats_lex_count, stats_lex_proportion, stats_unary_count, stats_unary_proportion, stats_nullary_count,
stats_nullary_proportion, stats_element_count, stats_element_proportion, stats_minmax_count, stats_minmax_proportion,
stats_occurrence_count, stats_occurrence_proportion, stats_multi_shared_vars, stats_edge_density, stats_Local_Variance,
standard_deviation_of_node_degree, normalised_standard_deviation_of_node_degree, clustering_coefficient, minimum_degree, normalised_minimum_degree,
maximum_degree, normalised_maximum_degree, median_degree, normalised_median_degree, mean_degree,
normalised_mean_degree, width_of_ordering, normalised_width_of_ordering, width_of_graph, normalised_width_of_graph,
SAC_literals, normalised_SAC_literals, stats_tightness_0, stats_tightness_25, stats_tightness_50,
stats_tightness_75, stats_tightness_100, stats_tightness_mean, stats_tightness_mean_normalised, stats_literal_tightness_0,
stats_literal_tightness_25, stats_literal_tightness_50, stats_literal_tightness_75, stats_literal_tightness_100, stats_literal_tightness_mean,
stats_literal_tightness_coeff_of_variation
2
to 2
, resulting in a PAR10 score of 6516.464
for the reduced model. Analogously, the model that was generated based on 3
of the originally 69
features resulted in a PAR10 score of 6415.798
.
Below, you can find the list of the selected features and solvers:
Selected Features:
stats_arity_50, stats_table_proportion, normalised_maximum_degree
Selected Solvers:
learning, standard