Documentation for RFSC_base
Base class for RFSC. Used to update RFSC parameters for DRFSC model
__init__(n_models=300, n_iters=150, tuning=50, tol=0.002, alpha=0.99, rip_cutoff=1, metric='roc_auc', verbose=False)
Parameters:
| Name | Type | Description | Default |
|---|---|---|---|
n_models |
int
|
Number of models generated per iteration. Default=300. |
300
|
n_iters |
int
|
Number of iterations. Default is 150. |
150
|
tuning |
float
|
Learning rate that dictates the speed of regressor inclusion probability (rip) convergence. Smaller values -> slower convergence. Default is 50. |
50
|
tol |
float
|
Tolerance condition. Default is 0.002. |
0.002
|
alpha |
float
|
Significance level for model pruning. Default is 0.99. |
0.99
|
rip_cutoff |
float
|
Determines rip threshold for feature inclusion in final model. Default=1. |
1
|
metric |
str
|
Optimization metric. Default='roc_auc'. Options: 'acc', 'roc_auc', 'weighted', 'avg_prec', 'f1', 'auprc'. |
'roc_auc'
|
verbose |
bool
|
Provides extra information. Defaults is False. |
False
|