tanu/ML_AI_training
1
0
Fork 0
Code Issues Pull requests Releases Wiki Activity

added all autosklearn informed hyperparams

Browse source
This commit is contained in:
Tanushree Tunstall 2022-05-26 04:51:47 +01:00
parent b5d29dd449
commit 8f8306d948
2 changed files with 141 additions and 7 deletions
Download patch file Download diff file Expand all files Collapse all files

129
classification_params_FS.py
View file

@ -7,8 +7,6 @@
# https://github.com/automl/auto-sklearn/tree/master/autosklearn/pipeline/components/classification
# TOADD:
# Extra Trees
https://github.com/automl/auto-sklearn/blob/master/autosklearn/pipeline/components/classification/extra_trees.py
# LDA
https://github.com/automl/auto-sklearn/blob/master/autosklearn/pipeline/components/classification/lda.py
# Multinomial_nb
@ -565,13 +563,138 @@ param_grid_svc = [
# 'clf': [SVC(**rs)],
# 'clf__kernel': ['poly', 'rbf', 'sigmoid']
, 'clf__kernel': ['rbf']
, 'clf__C' : [50, 10, 1.0, 0.1, 0.01]
# , 'clf__C' : [50, 10, 1.0, 0.1, 0.01]
, 'clf__C' : [1, 0.03, 10, 100, 1000, 10000, 32768]
, 'clf__gamma' : ['scale', 'auto']
}
]
#######################################################################
#######################################################################
#========
# LDA
# https://github.com/automl/auto-sklearn/blob/master/autosklearn/pipeline/components/classification/lda.py
#========
estimator = LinearDiscriminantAnalysis()
# Define pipleline with steps
pipe_lda = Pipeline([
('pre', MinMaxScaler())
, ('fs', RFECV(DecisionTreeClassifier(**rs), cv = cv, scoring = 'matthews_corrcoef'))
# , ('fs', RFECV(estimator, cv = cv, scoring = 'matthews_corrcoef'))
, ('clf', estimator)
])
# Define hyperparmeter space to search for
param_grid_lda = [
{
'fs__min_features_to_select' : [1,2]
# , 'fs__cv': [cv]
},
{
# 'clf': [LinearDiscriminantAnalysis()],
'clf__solver' : ['svd', 'lsqr', 'eigen'],
'clf__shrinkage' : [None, 'auto', 0, 0.5, 1],
}
]
#######################################################################
#========
# Multinomial_nb
# https://github.com/automl/auto-sklearn/blob/master/autosklearn/pipeline/components/classification/multinomial_nb.py
#========
estimator = MultinomialNB(**rs)
# Define pipleline with steps
pipe_mnb = Pipeline([
('pre', MinMaxScaler())
, ('fs', RFECV(DecisionTreeClassifier(), cv = cv, scoring = 'matthews_corrcoef'))
# , ('fs', RFECV(estimator, cv = cv, scoring = 'matthews_corrcoef'))
, ('clf', estimator)
])
# Define hyperparmeter space to search for
param_grid_mnb = [
{
'fs__min_features_to_select' : [1,2]
# , 'fs__cv': [cv]
},
{
# 'clf': [MultinomialNB()],
'clf__alpha': [0.01, 0.1, 1, 20, 25, 50, 55, 100]
}
]
#######################################################################
#========
# passive_aggressive
# https://github.com/automl/auto-sklearn/blob/master/autosklearn/pipeline/components/classification/passive_aggressive.py
#========
estimator = PassiveAggressiveClassifier(**rs, **njobs)
# Define pipleline with steps
pipe_pa = Pipeline([
('pre', MinMaxScaler())
, ('fs', RFECV(DecisionTreeClassifier(**rs), cv = cv, scoring = 'matthews_corrcoef'))
# , ('fs', RFECV(estimator, cv = cv, scoring = 'matthews_corrcoef'))
, ('clf', estimator)
])
# Define hyperparmeter space to search for
param_grid_pa = [
{
'fs__min_features_to_select' : [1,2]
# , 'fs__cv': [cv]
},
{
# 'clf': [PassiveAggressiveClassifier(**rs, **njobs)],
'clf__C' : [1, 0.03, 10, 100, 1000, 10000, 32768],
'clf__maxt_iter' : [1000, 500, 200, 100, 50, 10, 1],
'clf__loss' : ['hinge', 'squared_hinge'],
'clf_tol' : [1e-4, 1e-5, 1e-2, 1e-1]
}
]
#######################################################################
#========
# SGD
# https://github.com/automl/auto-sklearn/blob/master/autosklearn/pipeline/components/classification/sgd.py
#========
estimator = SGDClassifier(**rs, **njobs)
# Define pipleline with steps
pipe_sgd = Pipeline([
('pre', MinMaxScaler())
, ('fs', RFECV(DecisionTreeClassifier(**rs), cv = cv, scoring = 'matthews_corrcoef'))
# , ('fs', RFECV(estimator, cv = cv, scoring = 'matthews_corrcoef'))
, ('clf', estimator)
])
# Define hyperparmeter space to search for
param_grid_sgd = [
{
'fs__min_features_to_select' : [1,2]
# , 'fs__cv': [cv]
},
{
# 'clf': [SGDClassifier(**rs, **njobs)],
'clf__loss': = ['hinge', 'log', 'modified_huber', 'squared_hinge', 'perceptron'],
'clf__penalty':['l1', 'l2', 'elasticnet'],
'clf__alpha': [0.0000001, 0.00001, 0.0001, 0.01, 0.1, 1, 10, 100], #autosk learn: 1e-7, 1e-1, log=True, default_value=0.0001
'clf__learning_rate': ['constant', 'optimal', 'invscaling', 'adaptive'],
'clf__eta0' : [0.0000001, 0.00001, 0.0001, 0.01] # autosklearn 1e-7, 1e-1, default_value=0.01, log=True
}
]