-
Notifications
You must be signed in to change notification settings - Fork 0
/
adaboost.py
166 lines (140 loc) · 5.82 KB
/
adaboost.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
import numpy as np
import pandas as pd
import os
import sys
from os import path as op
from sklearn.ensemble import AdaBoostClassifier
from sklearn.model_selection import train_test_split
from sklearn.metrics import accuracy_score
import time
from matplotlib import pyplot as plt
from sklearn.tree import DecisionTreeClassifier
from scipy import stats
from tqdm import tqdm
def mode(x):
if (len(x)==0):
return -1
return stats.mode(x.flatten())[0][0]
class WeightedStump():
def __init__(self):
pass
def fit(self, X, y, sample_weight):
if np.unique(y).size <= 1:
return
N, D = X.shape
w = sample_weight
pos_weights = w[y==1]
neg_weights = w[y==-1]
class_vals, count = np.unique(y, return_counts=True)
self.stump_positive_label = class_vals[np.argmax(count)]
self.stump_negative_label = None
self.stump_feature = None
self.stump_threshold = None
X = np.round(X)
max_ig = float("-inf")
for d in np.arange(D):
for value in np.unique(X):
positive_label_candidate = mode(y[X[:,d] > value])
negative_label_candidate = mode(y[X[:,d] <= value])
y_pred = positive_label_candidate * np.ones(N)
y_pred[X[:, d] <= value] = negative_label_candidate
# Handle the empty list condition and skip this.
try:
positive_rate = np.unique(y[X[:,d] > value], return_counts=True)[1][1] / N
except IndexError:
continue
negative_rate = 1 - positive_rate
y2 = np.asarray([np.sum(neg_weights), np.sum(pos_weights)]) / np.sum(w)
sub_y_yes = y[X[:,d] > value]
sub_w_yes = w[X[:,d] > value]
sub_positive_rate = np.asarray([np.sum(sub_w_yes[sub_y_yes == -1]), np.sum(sub_w_yes[sub_y_yes == 1])])/np.sum(w)
sub_negative_rate = 1 - sub_positive_rate
ig = (stats.entropy(y2)) - ((positive_rate *
stats.entropy(sub_positive_rate))) - ((negative_rate * stats.entropy(sub_negative_rate)))
if ig > max_ig:
max_ig = ig
self.stump_feature = d
self.stump_threshold = value
self.stump_positive_label = positive_label_candidate
self.stump_negative_label = negative_label_candidate
return
def predict(self, X):
M, _ = X.shape
X = np.round(X)
if self.stump_feature is None:
exit("Error! None value stump_feature encountered!")
y_pred = np.zeros(M)
y_pred[X[:, self.stump_feature] <= self.stump_threshold] = self.stump_negative_label
y_pred[X[:, self.stump_feature] > self.stump_threshold] = self.stump_positive_label
return y_pred
class AdaBoost():
def __init__(self, n_estimators = 10):
self.n_estimators = n_estimators
self.estimators= None
self.alphas = None
def fit(self, X, y):
n_samples, n_features = np.shape(X)
w = np.full(n_samples, (1 / n_samples))
self.estimators = []
self.alphas = []
for _ in range(self.n_estimators):
sklearn_stump = WeightedStump()
sklearn_stump.fit(X,y,w)
predictions = sklearn_stump.predict(X)
accuracy = accuracy_score(y, predictions, sample_weight=w)
error = 1 - accuracy
alpha = 0.5 * np.log((1.0 - error) / (error + 1e-10))
self.alphas.append(alpha)
w *= np.exp(-alpha * y * predictions)
w /= np.sum(w)
self.estimators.append(sklearn_stump)
def predict(self, X):
n_samples = X.shape[0]
y_pred = np.zeros((n_samples, 1))
for i, clf in enumerate(self.estimators):
predictions = np.expand_dims(clf.predict(X),1)
y_pred += self.alphas[i] * predictions
# Return sign of prediction sum
y_pred = np.sign(y_pred).flatten()
return y_pred
def score(self, X, y):
preds = self.predict(X)
labels = np.squeeze(y)
assert preds.shape == labels.shape
return np.mean(preds == labels)
def adaboost(dataset:str) -> None:
# Read the data and replace question mark values with nans and then impute nans with ones.
df = pd.read_csv(dataset, sep=',' , header=None).replace('?', np.nan).fillna(value=1).astype('int32')
y = df[10].to_numpy()
# Encode labels.
y[y==2] = 1
y[y==4] = -1
X = df[[1, 2, 3, 4, 5, 6, 7, 8, 9]].to_numpy().astype('int32')
X_train, X_test, y_train, y_test = train_test_split(
X, y, test_size=0.2, random_state=42)
# Adaboost sanity check
# sk_clf = AdaBoostClassifier(base_estimator=DecisionTreeClassifier(max_depth=1), n_estimators=100)
# sk_clf.fit(X_train,y_train)
# print(sk_clf.score(X_test,y_test))
tr_errors = []
test_errors = []
for n_estimators in tqdm(np.arange(1, 101)):
model = AdaBoost(n_estimators=n_estimators)
model.fit(X_train, y_train)
tr_error = 1.0-model.score(X_train, y_train)
test_error= 1.0-model.score(X_test, y_test)
tr_errors.append(tr_error)
test_errors.append(test_error)
print("Training error with", n_estimators, "estimators:",tr_error)
print("Test error with", n_estimators, "estimators:",test_error)
plt.plot(np.arange(1, 101), tr_errors, 'b-',
np.arange(1, 101), test_errors, 'r-')
plt.title("AdaBoost training and test errors")
plt.xlabel("number of estimators")
plt.ylabel("Error")
plt.legend(['training error', 'test error'])
plt.show()
plt.clf()
return
if __name__ == "__main__":
adaboost("data/breast-cancer-wisconsin.data")