-
Notifications
You must be signed in to change notification settings - Fork 2
/
Copy pathmodel_training.py
178 lines (133 loc) · 7.62 KB
/
model_training.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
from sklearn.preprocessing import LabelBinarizer
from sklearn.model_selection import GridSearchCV, StratifiedKFold
from sklearn.tree import DecisionTreeClassifier
from sklearn.ensemble import GradientBoostingClassifier, RandomForestClassifier
from sklearn.linear_model import LogisticRegression
from sklearn.neighbors import KNeighborsClassifier
from sklearn.model_selection import train_test_split
from sklearn.svm import SVC
from sklearn.metrics import f1_score, recall_score, precision_recall_curve
from sklearn.decomposition import PCA
from sklearn.cluster import KMeans
import pickle
from ipywidgets import IntProgress
from mlxtend.plotting import plot_learning_curves, plot_decision_regions
from mlxtend.plotting import plot_confusion_matrix
from keras.models import Sequential
from keras.layers import Dense, Conv1D, MaxPooling1D, Flatten
from keras.wrappers.scikit_learn import KerasClassifier
from keras.optimizers import RMSprop
from keras_tqdm import TQDMNotebookCallback
from youtube_audioset import get_data, get_recursive_sound_names, get_all_sound_names
from youtube_audioset import explosion_sounds, motor_sounds, wood_sounds, human_sounds, nature_sounds
ambient_sounds, impact_sounds = get_all_sound_names()
explosion_sounds = get_recursive_sound_names(explosion_sounds)
motor_sounds = get_recursive_sound_names(motor_sounds)
wood_sounds = get_recursive_sound_names(wood_sounds)
human_sounds = get_recursive_sound_names(human_sounds)
nature_sounds = get_recursive_sound_names(nature_sounds)
#Read the balanced data created by running the balancing_datasets.py
with open('balanced_data.pkl','rb') as f:
df=pickle.load(f)
print(df.shape)
df['labels']=df['labels_name']
# Binarize the labels
name_bin = LabelBinarizer().fit(ambient_sounds + impact_sounds)
labels_split = df['labels'].apply(pd.Series).fillna('None')
labels_binarized = name_bin.transform(labels_split[labels_split.columns[0]])
for column in labels_split.columns:
labels_binarized |= name_bin.transform(labels_split[column])
labels_binarized = pd.DataFrame(labels_binarized, columns = name_bin.classes_)
print(labels_binarized.shape)
print df.shape[0], "examples"
print "Percentage Impact Sounds:", (labels_binarized[impact_sounds].sum(axis=1) > 0).mean()
print "Percentage Ambient Sounds:", (labels_binarized[ambient_sounds].sum(axis=1) > 0).mean()
print(labels_binarized.mean())
df_filtered = df.loc[df.features.apply(lambda x: x.shape[0] == 10)]
# df_filtered = df.loc[df['labels'].apply(lambda x: (len(x) == 1)) & df.features.apply(lambda x: x.shape[0] == 10)]
labels_filtered = labels_binarized.loc[df_filtered.index,:]
df_train, df_test, labels_binarized_train, labels_binarized_test = train_test_split(df_filtered, labels_filtered,
test_size=0.33, random_state=42,
stratify=labels_filtered.any(axis=1)*1)
X_train = np.array(df_train.features.apply(lambda x: x.flatten()).tolist())
X_train_standardized = X_train / 255
X_test = np.array(df_test.features.apply(lambda x: x.flatten()).tolist())
X_test_standardized = X_test / 255
y_train = (labels_binarized_train[impact_sounds].any(axis=1)*1).values
y_test = (labels_binarized_test[impact_sounds].any(axis=1)*1).values
print labels_filtered.loc[:,explosion_sounds].any(axis=1).mean()
print labels_filtered.loc[:,motor_sounds].any(axis=1).mean()
print labels_filtered.loc[:,wood_sounds].any(axis=1).mean()
print labels_filtered.loc[:,human_sounds].any(axis=1).mean()
print labels_filtered.loc[:,nature_sounds].any(axis=1).mean()
labels_filtered.loc[:,impact_sounds].any(axis=1).mean()
clf1_ = LogisticRegression(max_iter=1000)
# clf1_train = X_train.reshape((-1,10,128)).max(axis=1)
# clf1_test = X_test.reshape((-1,10,128)).max(axis=1)
clf1_train = X_train
clf1_test = X_test
# clf1_train_target = labels_binarized_train.loc[:,motor_sounds].any(axis=1)
# clf1_test_target = labels_binarized_test.loc[:,motor_sounds].any(axis=1)
clf1_train_target = labels_binarized_train.loc[:,impact_sounds].any(axis=1)
clf1_test_target = labels_binarized_test.loc[:,impact_sounds].any(axis=1)
clf1_.fit(clf1_train, clf1_train_target)
clf1_train_prediction = clf1_.predict(clf1_train)
clf1_test_prediction = clf1_.predict(clf1_test)
clf1_test_prediction_prob = clf1_.predict_proba(clf1_test)[:,1]
clf1_conf_train_mat = pd.crosstab(clf1_train_target, clf1_train_prediction, margins=True)
print('Train precsion and recall for Logistic regression')
print('=============================================')
print "Train Precision:", clf1_conf_train_mat[True][True] / float(clf1_conf_train_mat[True]['All'])
print "Train Recall:", clf1_conf_train_mat[True][True] / float(clf1_conf_train_mat['All'][True])
print "Train Accuracy:", (clf1_train_prediction == clf1_train_target).mean()
print(clf1_conf_train_mat)
clf1_conf_test_mat = pd.crosstab(clf1_test_target, clf1_test_prediction, margins=True)
print('Test precsion and recall for Logistic regression')
print('=============================================')
print "Test Precision:", clf1_conf_test_mat[True][True] / float(clf1_conf_test_mat[True]['All'])
print "Test Recall:", clf1_conf_test_mat[True][True] / float(clf1_conf_test_mat['All'][True])
print "Test Accuracy:", (clf1_test_prediction == clf1_test_target).mean()
print(clf1_conf_test_mat)
def create_keras_model():
# create model
model = Sequential()
model.add(Conv1D(40, input_shape=(1280,1), kernel_size=128, strides=128, activation='relu', padding='same'))
model.add(Conv1D(100, kernel_size=3, activation='relu', padding='same'))
model.add(Dense(10, activation='relu'))
model.add(Dense(1, activation='sigmoid'))
model.add(MaxPooling1D(10))
model.add(Flatten())
print model.summary()
# Compile model
model.compile(loss='binary_crossentropy', optimizer=RMSprop(lr=5e-4), metrics=['accuracy'])
return model
clf2_train = X_train.reshape((-1,1280,1))
clf2_test = X_test.reshape((-1,1280,1))
clf2_train_target = labels_binarized_train.loc[:,impact_sounds].any(axis=1)
clf2_test_target = labels_binarized_test.loc[:,impact_sounds].any(axis=1)
model = create_keras_model()
model_traing=model.fit(clf2_train, clf2_train_target,
epochs=500, batch_size=5000, verbose=False,
validation_data = (clf2_test, clf2_test_target))
clf2_train_prediction = model.predict(clf2_train).ravel().round()
clf2_test_prediction = model.predict(clf2_test).ravel().round()
clf2_test_prediction_prob = model.predict(clf2_test).ravel()
print "Train Accuracy:", (clf2_train_prediction == clf2_train_target).mean()
print "Test Accuracy:", (clf2_test_prediction == clf2_test_target).mean()
clf2_conf_train_mat = pd.crosstab(clf2_train_target, clf2_train_prediction, margins=True)
print("Training Precision and recall for Keras model")
print('=============================================')
print "Train Precision:", clf2_conf_train_mat[True][True] / float(clf2_conf_train_mat[True]['All'])
print "Train Recall:", clf2_conf_train_mat[True][True] / float(clf2_conf_train_mat['All'][True])
print "Train Accuracy:", (clf2_train_prediction == clf2_train_target).mean()
print(clf2_conf_train_mat)
clf2_conf_test_mat = pd.crosstab(clf2_test_target, clf2_test_prediction, margins=True)
print("Testing Precision and recall for Keras model")
print('=============================================')
print "Test Precision:", clf2_conf_test_mat[True][True] / float(clf2_conf_test_mat[True]['All'])
print "Test Recall:", clf2_conf_test_mat[True][True] / float(clf2_conf_test_mat['All'][True])
print "Test Accuracy:", (clf2_test_prediction == clf2_test_target).mean()
print(clf2_conf_test_mat)