Andrew Ng Coursera week2 Logistic Regression with a Neural Network mindset 程式設計作業代碼祥注與資料集

這是coursea吳恩達老師的Neural Networks and Deep Learning week2的程式設計作業和離線資料

import numpy as np
import matplotlib.pyplot as plt 
import h5py
import scipy
from PIL import Image 
from scipy import ndimage
from lr_utils import load_dataset

#讀入資料庫，如圖像，訓練集，測試集
train_set_x_orig,train_set_y,test_set_x_orig,test_set_y,classes=load_dataset()


#此乃訓練集數量
m_train=train_set_x_orig.shape[0]
#此乃測試集數量
m_test=train_set_x_orig.shape[0]
#此乃圖像的點陣規模
num_px=train_set_x_orig.shape[1]
#将矩陣标準化
train_set_x_flatten=train_set_x_orig.reshape(train_set_x_orig.shape[0],-1).T
test_set_x_flatten=test_set_x_orig.reshape(test_set_x_orig.shape[0],-1).T

train_set_x=train_set_x_flatten/255
test_set_x=test_set_x_flatten/255

#sigmoid函數，亦稱為邏輯回歸函數
def sigmoid(z):
	s=1/(1+np.exp(-z))
	return s
#print("sigmoid"+str(sigmoid(np.array([0,2]))))


def init(dim):
	w=np.zeros((dim,1))
	b=0
	return w,b

#前向傳播
def pro(w,b,X,Y):
	m=X.shape[1]
	A=sigmoid(np.dot(w.T ,X)+b)
	#損失函數
	cost=(-1/m)*np.sum(Y*np.log(A)+(1-Y)*np.log(1-A))
	#此處記錄梯度
	dw=(1/m)*np.dot(X,(A-Y).T)
	db=(1/m)*np.sum(A-Y)
	cost=np.squeeze(cost)
	#梯度cache
	grads={"dw":dw,"db":db}

	return grads,cost
#反向傳播和梯度下降
def opt(w,b,X,Y,num_ite,learning_rate,print_cost=False):
	costs=[]
	#不斷疊代進行梯度下降
	for i in range(num_ite):
		grads,cost=pro(w,b,X,Y)
		dw=grads["dw"]
		db=grads["db"]
		w=w-learning_rate*dw 
		b=b-learning_rate*db 
		#每100次疊代輸出損失函數作為對比
		if i%100==0:
			costs.append(cost)
		if print_cost and i%100==0:
			print("Cost after iteration %i:%f"%(i,cost))

	params={"w":w,"b":b}
	grads={"dw":dw,"db":db}
	return params,grads,costs 
#預測函數
def predict(w,b,X):
	m=X.shape[1]
	Y_prediction=np.zeros((1,m))
	w=w.reshape(X.shape[0],1)

	A=sigmoid(np.dot(w.T,X)+b)
#二進制分類
	for i in range(A.shape[1]):
		if A[0,i]>0.5 :
			Y_prediction[0,i]=1
		else:
			Y_prediction[0,i]=0

	return Y_prediction

def model(X_train,Y_train,X_test,Y_test,num_ite=2000,learning_rate=0.5,print_cost=True):
	w,b=init(X_train.shape[0])

	params,gards,costs=opt(w,b,X_train,Y_train,num_ite,learning_rate,print_cost=True)

	w=params["w"]
	b=params["b"]

	Y_prediction_test=predict(w,b,X_test)
	Y_prediction_train=predict(w,b,X_train)

	print("train accuracy: {} %".format(100 - np.mean(np.abs(Y_prediction_train - Y_train)) * 100))
	print("test accuracy: {} %".format(100 - np.mean(np.abs(Y_prediction_test - Y_test)) * 100))

	d={"cost":costs,"Y_prediction_train":Y_prediction_train,"Y_prediction_test":Y_prediction_test,"w":w,"b":b,"learning_rate":learning_rate,"num_ite":num_ite}

	return d

d=model(train_set_x,train_set_y,test_set_x,test_set_y,num_ite=2000,learning_rate=0.005,print_cost=True)

index = 25
plt.imshow(test_set_x[:,index].reshape((num_px, num_px, 3)))
plt.show()
print ("y = " + str(test_set_y[0,index]) + ", you predicted that  ")

if d["Y_prediction_test"][0,index]==1:
	print("it's a cat.")
else :
	print("it'not a cat")
           

https://github.com/ZeroT900sk/week2