수수께끼 고양이

딥러닝 7 공식 적용하기 from math import exp x1,x2=0.05,0.10 w1,w2=0.15,0.20 w3,w4=0.25,0.30 b1,b2=0.35,0.35 w5,w6=0.40,0.45 w7,w8=0.50,0.55 b3,b4=0.60,0.60 y1T,y2T=0.01,0.99 lr=0.01 EPOCH=1000 for epoch in range(EPOCH): h1=x1*w1+x2*w2+1*b1 h2=x1*w3+x2*w4+1*b2 # ReLU feed forward h1=h1 if h1>0 else 0 h2=h2 if h2>0 else 0 y1=h1*w5+h2*w6+1*b3 y2=h1*w7+h2*w8+1*b4 # sigmoid feed forward y1=1/(1+exp(-y1)) y2=1..

1. sigmoid 함수 그려보기(계단 함수, 역전파) import numpy as np def sigmoid(x): return 1/(1+np.exp(-x)) # x=np.random.uniform(-10,10,1000) x=np.random.uniform(-10,100,10000) y=sigmoid(x) import matplotlib.pyplot as plt plt.plot(x,y,'r.') plt.show() 2. ReLU 함수 그려보기 import numpy as np def ReLU(x): return x*(x>0) x=np.random.uniform(-10,10,1000) y=ReLU(x) import matplotlib.pyplot as plt plt.plot(x,y,'g.') plt.sh..

numpy, matplotlib 라이브러리를 이용하여 딥러닝 학습 예제 1. w, b, E의 관계 살펴보기 import numpy as np # numpy 행렬 import matplotlib.pyplot as plt # matplotlib 그래프로 표현해줌 fig = plt.figure(figsize=(8,8)) ax = fig.add_subplot(projection='3d') ax.set_title("wbE",size=20) ax.set_xlabel("w", size=14) ax.set_xlabel("b", size=14) ax.set_xlabel("E", size=14) x=2 yT=10 w=np.random.uniform(-200,200,10000) b=np.random.uniform(-200,..

연습문제 1. 2입력 2은닉 2출력 인공신경망 x1,x2=0.05,0.10 w1,w2=0.15,0.20 w3,w4=0.25,0.30 b1,b2=0.35,0.35 w5,w6=0.40,0.45 w7,w8=0.50,0.55 b3,b4=0.60,0.60 y1T,y2T=0.01,0.99 lr=0.01 for epoch in range(2000): h1=x1*w1+x2*w2+1*b1 h2=x1*w3+x2*w4+1*b2 y1=h1*w5+h2*w6+1*b3 y2=h1*w7+h2*w8+1*b4 E=((y1-y1T)**2+(y2-y2T)**2)/2 y1E=y1-y1T y2E=y2-y2T w5E=y1E*h1 w6E=y1E*h2 w7E=y2E*h1 w8E=y2E*h2 b3E=y1E*1 b4E=y2E*1 h1E=y1E*w5+y..

연습문제 1. 2입력 3출력 인공신경망을 구현하고 학습하기 공식 정리 # 순전파 y1=x1*w1+x2*w2+1*b1 y2=x1*w3+x2*w4+1*b2 y3=x1*w5+x2*w6+1*b3 # 평균 제곱 오차 E=(y1-y1T)**2/2+(y2-y2T)**2/2+(y3-y3T)**2/2 # 역전파 오차 y1E=y1-y1T y2E=y2-y2T y3E=y3-y3T # 입력 역전파 x1E=y1E*w1+y2E*w3+y3E*w5 x2E=y1E*w2+y2E*w4+y2E*w6 #가중치, 편향 순전파 y1=x1*w1+x2*w2+1*b1 y2=x1*w3+x2*w4+1*b2 y3=x1*w5+x2*w6+1*b3 y1=w1*x1+w2*x2+b1*1 y2=w3*x1+w4*x2+b2*1 y3=w5*x1+w6*x2+b3*1 # 가중..

1. 순전파 : y1=x1*w1+x2*w2+1*b1, y2=x1*w3+x2*w4+1*b2 x1,x2=2,3 w1,w2=3,4 w3,w4=5,6 b1,b2=1,2 y1=x1*w1+x2*w2+1*b1 y2=x1*w3+x2*w4+1*b2 (y1,y2) 2. 평균 제곱 오차 : E=(y1-y1T)*(y1-y1T)/2+(y2-y2T)*(y2-y2T)/2 y1T,y2T=27,-30 E=(y1-y1T)**2/2+(y2-y2T)**2/2 E 3. 역전파 오차 : y1E=y1-y1T, y2E=y2-y2T y1E=y1-y1T y2E=y2-y2T (y1E,y2E) 4. 입력 역전파 : x1E=y1E*w1+y2E*w3, x2E=y1E*w2+y2E*w4 5. 가중치, 편향 순전파 y1=x1*w1+x2*w2+1*b1 y2=x1*..