binaryclassification
import numpy as np import matplotlib.pyplot as plt import pandas as pd import sklearn import sklearn.preprocessing as pre
df=pd.read_csv('data\glassi\glass.data') df.head()
id | RI | Na | Mg | Al | Si | K | Ca | Ba | Fe | class | |
---|---|---|---|---|---|---|---|---|---|---|---|
0 | 1 | 1.52101 | 13.64 | 4.49 | 1.10 | 71.78 | 0.06 | 8.75 | 0.0 | 0.0 | 1 |
1 | 2 | 1.51761 | 13.89 | 3.60 | 1.36 | 72.73 | 0.48 | 7.83 | 0.0 | 0.0 | 1 |
2 | 3 | 1.51618 | 13.53 | 3.55 | 1.54 | 72.99 | 0.39 | 7.78 | 0.0 | 0.0 | 1 |
3 | 4 | 1.51766 | 13.21 | 3.69 | 1.29 | 72.61 | 0.57 | 8.22 | 0.0 | 0.0 | 1 |
4 | 5 | 1.51742 | 13.27 | 3.62 | 1.24 | 73.08 | 0.55 | 8.07 | 0.0 | 0.0 | 1 |
X,y=df.iloc[:,1:-1],df.iloc[:,-1]
X,y=np.array(X),np.array(y)
#change the value the element
for idx,class_name in enumerate(sorted(list(set(y)))):
y[y==class_name]=idx
y
array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 4, 4, 4, 4, 4, 4, 4, 4, 4, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5], dtype=int64)
#make the matrix's elements 2 value #if element doesn't equals to 1 then make it 0 #'1' stands for the '2' class for i in range(len(y)): if y[i]!=1: y[i]=0
#split our training dataset randomly from sklearn.model_selection import train_test_split X_train,X_test,y_train,y_test=train_test_split(X,y,test_size=0.15,random_state=44)
X_train.shape,y_train.shape,X_test.shape,y_test.shape
((181, 9), (181,), (33, 9), (33,))
f_mean=np.mean(X_train,axis=0) f_std=np.std(X_train,axis=0) f_mean,f_std
(array([1.51832884e+00, 1.33736464e+01, 2.69287293e+00, 1.46425414e+00, 7.26391160e+01, 5.17016575e-01, 8.95314917e+00, 1.71104972e-01, 6.02762431e-02]), array([0.00300427, 0.79769555, 1.42353328, 0.49169919, 0.77056863, 0.69105168, 1.42892902, 0.5002639 , 0.10131419]))
#standardize training set X_train=(X_train-f_mean)/f_std X_test=(X_test-f_mean)/f_std
theta = np.zeros((X_train.shape[1] + 1)) theta.shape
(10,)
#add constant parameter X_train = np.concatenate((np.ones((X_train.shape[0], 1)), X_train), axis=1) X_test = np.concatenate((np.ones((X_test.shape[0], 1)), X_test), axis=1) X_train.shape,X_test.shape,theta.shape
((181, 10), (33, 10), (10,))
#initialize the parameter np.random.seed(42) theta = np.random.rand(*theta.shape) theta
array([0.37454012, 0.95071431, 0.73199394, 0.59865848, 0.15601864, 0.15599452, 0.05808361, 0.86617615, 0.60111501, 0.70807258])
#cross_entropy_loss: loss function #h: hypothesis function #gradient: gradient function num_epoch=500000 for epoch in range(num_epoch): logist = np.dot(X_train, theta) h = 1 / (1 + np.exp(-logist)) cross_entropy_loss = (-y_train * np.log(h) - (1 - y_train) * np.log(1 - h)).mean() gradient = np.dot((h - y_train), X_train) / y_train.size theta = theta - 0.01*gradient if epoch%100000==0: print('Epoch={}\tLoss={}'.format(epoch,cross_entropy_loss))
Epoch=0 Loss=0.9770836920534414 Epoch=100000 Loss=0.5884129057196792 Epoch=200000 Loss=0.5828823869347305 Epoch=300000 Loss=0.5798937167992417 Epoch=400000 Loss=0.5782071252958373
h_test = 1 / (1 + np.exp(-np.dot(X_test, theta))) #accurancy ((h_test > 0.5) == y_test).sum() / y_test.size
0.8484848484848485