def DTree(file1):
# Read Data
data = Data()
data_train, data_test, output_train, output_test, unscaled_data_test, activities_labels, feature_labels, groups_at_training, groups_at_testing = data
InputData(data_train, data_test, output_test, name='Initial Test Data')
spl = GroupShuffleSplit(n_splits=6, test_size=0.3, random_state=0)
classifier = DecisionTreeClassifier()
classifier_name = DecisionTreeClassifier.__name__
name = "DTree"
params_grid = [{
'criterion': ["gini", "entropy"],
'max_leaf_nodes': [30, 50, 70, 100]
}]
classifier = CrossValidationWithGridSearch(spl, classifier, data_train,
output_train,
groups_at_training, params_grid,
classifier_name, file1)
pred_data_test, pred_data_train = model_training(classifier, data_train,
output_train, data_test,
output_test)
Results(data_train, output_train, pred_data_train, output_test,
pred_data_test, activities_labels, data_test, classifier, file1,
name)
def RandomForest(file1):
# Read Data
data = Data()
data_train, data_test, output_train, output_test, unscaled_data_test, activities_labels, feature_labels, groups_at_training, groups_at_testing = data
InputData(data_train, data_test, output_test, name='Initial Test Data')
spl = GroupShuffleSplit(n_splits=6, test_size=0.3, random_state=0)
classifier = RandomForestClassifier(n_estimators=100,
criterion='gini',
max_depth=70,
random_state=50)
fig, ax = plt.subplots()
name = "Random Forest"
alg_name = RandomForestClassifier.__name__
CrossValidationWithoutGridSearch(spl,
classifier,
data_train,
output_train,
groups_at_training,
alg_name,
file1,
ax,
n_splits=6,
lw=10)
pred_data_test, pred_data_train = model_training(classifier, data_train,
output_train, data_test,
output_test)
Results(data_train, output_train, pred_data_train, output_test,
pred_data_test, activities_labels, data_test, classifier, file1,
name)
def LR_model(file1):
# Read Data
data = Data()
data_train, data_test, output_train, output_test, unscaled_data_test, activities_labels, feature_labels, groups_at_training, groups_at_testing = data
InputData(data_train, data_test, output_test, name='Initial Test Data')
spl = GroupShuffleSplit(n_splits=6, test_size=0.3, random_state=0)
#spl = GroupShuffleSplit(n_splits=4,train_size=.6, test_size=.4, random_state=42)
classifier = LogisticRegression(tol=1e-3, penalty='l2')
classifier_name = LogisticRegression.__name__
name = "LR"
params_grid = [{
'solver': ['newton-cg'],
'C': [0.5, 1, 10, 100, 1000],
'max_iter': [10000, 100000]
}, {
'solver': ['lbfgs'],
'C': [0.5, 1, 10, 100, 1000],
'max_iter': [10000, 100000]
}, {
'solver': ['liblinear'],
'C': [0.5, 1, 10, 100, 1000],
'max_iter': [10000, 100000]
}, {
'solver': ['sag'],
'C': [0.5, 1, 10, 100, 1000],
'max_iter': [10000, 100000]
}, {
'solver': ['saga'],
'C': [0.5, 1, 10, 100, 1000],
'max_iter': [10000, 100000]
}]
classifier = CrossValidationWithGridSearch(spl, classifier, data_train,
output_train,
groups_at_training, params_grid,
classifier_name, file1)
pred_data_test, pred_data_train = model_training(classifier, data_train,
output_train, data_test,
output_test)
Results(data_train, output_train, pred_data_train, output_test,
pred_data_test, activities_labels, data_test, classifier, file1,
name)
def LR_model(file1):
# Read Data
data = Data()
data_train, data_test, output_train, output_test, unscaled_data_test, activities_labels, feature_labels, groups_at_training, groups_at_testing = data
InputData(data_train, data_test, output_test, name='Initial Test Data')
spl = GroupShuffleSplit(n_splits=6, test_size=0.3, random_state=0)
#spl = GroupShuffleSplit(n_splits=10,train_size=.6, test_size=.4, random_state=42)
classifier = SVC(kernel='rbf', gamma=0.01, C=100, degree=3)
classifier_name = SVC.__name__
name = "SVC"
params_grid = [{
'kernel': ['rbf'],
'gamma': [1e-2, 1e-1, 1, 2],
'C': [0.1, 1, 10, 50, 100, 150]
}, {
'kernel': ['linear'],
'C': [0.1, 1, 10, 50, 100, 150]
}, {
'kernel': ['poly'],
'gamma': [1e-2, 1e-1, 1, 2],
'degree': [2, 3, 4, 5],
'C': [0.1, 1, 10, 50, 100, 150]
}, {
'kernel': ['sigmoid'],
'gamma': [1e-2, 1e-1, 1, 2],
'C': [0.1, 1, 10, 50, 100, 150]
}]
#classifier = CrossValidationWithGridSearch(spl,classifier,data_train,output_train,groups_at_training,params_grid,classifier_name,file1)
pred_data_test, pred_data_train = model_training(classifier, data_train,
output_train, data_test,
output_test,
groups_at_training)
Results(data_train, output_train, pred_data_train, output_test,
pred_data_test, activities_labels, data_test, classifier, file1,
name)
def LDA_model(file1):
# Read Data
data = Data()
data_train, data_test, output_train, output_test, unscaled_data_test, activities_labels, feature_labels, groups_at_training, groups_at_testing = data
num_training = cound_groups(groups_at_training)
num_test = cound_groups(groups_at_testing)
print(num_training, num_test)
PlotTheInputData(output_train, activities_labels, "Train Data")
PlotTheInputData(output_test, activities_labels, 'Test Data')
InputData(data_train, data_test, output_test, name='Initial Test Data')
#lpl = LeavePGroupsOut(n_groups=2)
spl = GroupShuffleSplit(n_splits=6, test_size=0.3, random_state=0)
classifier = LinearDiscriminantAnalysis()
fig, ax = plt.subplots()
name = "LDA"
alg_name = LinearDiscriminantAnalysis.__name__
CrossValidationWithoutGridSearch(spl,
classifier,
data_train,
output_train,
groups_at_training,
alg_name,
file1,
ax,
n_splits=6,
lw=10)
pred_data_test, pred_data_train = model_training(classifier, data_train,
output_train, data_test,
output_test)
Results(data_train, output_train, pred_data_train, output_test,
pred_data_test, activities_labels, data_test, classifier, file1,
name)
# Compare Algorithms
import pandas
import numpy as np
import matplotlib.pyplot as plt
from Data.Select_From_model import Data
from sklearn import model_selection
from sklearn.linear_model import LogisticRegression
from sklearn.tree import DecisionTreeClassifier
from sklearn.discriminant_analysis import LinearDiscriminantAnalysis
from sklearn.svm import SVC
from sklearn.ensemble import RandomForestClassifier
from sklearn.metrics import accuracy_score
from UsefullFunctions import Add_Noise
from sklearn.preprocessing import MinMaxScaler
data = Data()
data_train, data_test, output_train, output_test, unscaled_data_test, activities_labels, feature_labels, groups_at_training, groups_at_testing = data
# prepare configuration for cross validation test harness
seed = 7
# prepare models
models = []
models.append(
('LR', LogisticRegression(C=10, solver='liblinear', max_iter=10000)))
models.append(('LDA', LinearDiscriminantAnalysis()))
models.append(('RF',
RandomForestClassifier(n_estimators=1000,
criterion='gini',
max_depth=20,
random_state=50)))
models.append(
('DTree', DecisionTreeClassifier(criterion='entropy', max_leaf_nodes=70)))