内容来自 麦子学院 深度学习基础及介绍
1 神经网络
多层向前神经网络:Multilayer Feed-Forward Neural Network
定义:有输入层、隐藏层、输出层,每层由单元组成。输入层由训练集的特征向量传入,经过单元的权重加权求和,作为下一层的输入。如果有足够多的隐藏层,和足够大的训练集,可以模拟出任何方程。
设计结构:确定层数、每层单元数;输入特征向量先要标准化到0~1之间;可以编码离散变量;可用于分类或回归;隐藏层的多少则没有规定,根据经验选取;
交叉验证:在给定的建模样本中,拿出大部分样本进行建模型,留小部分样本用刚建立的模型进行预报,并求这小部分样本的预报误差,记录它们的平方加和。这个过程一直进行,直到所有的样本都被预报了一次而且仅被预报一次。把每个样本的预报误差平方加和,称为PRESS
核心算法:backpropagation:对比预测值与真实值之间的值,反向传播,最小化误差,更新每个连接的权重。初始权重和偏向随机分配。
下面的程序定义了用于神经网络算法的类:
import numpy as np #科学计算库
def tanh(x):
return np.tanh(x)
def tanh_deriv(x):
return 1.0 - np.tanh(x)*np.tanh(x)
def logistic(x):
return 1/(1 + np.exp(-x))
def logistic_derivative(x):
return logistic(x)*(1-logistic(x))
class NeuralNetwork:
def __init__(self, layers, activation='tanh'):
"""
:param layers: A list containing the number of units in each layer.
Should be at least two values
:param activation: The activation function to be used. Can be
"logistic" or "tanh"
"""
if activation == 'logistic':
self.activation = logistic
self.activation_deriv = logistic_derivative
elif activation == 'tanh':
self.activation = tanh
self.activation_deriv = tanh_deriv
self.weights = [] #存放权重,初始化时随机产生
for i in range(1, len(layers) - 1): #除输出层外均需要初始化
self.weights.append((2*np.random.random((layers[i - 1] + 1, layers[i] + 1))-1)*0.25)
self.weights.append((2*np.random.random((layers[i] + 1, layers[i + 1]))-1)*0.25)
#############################################################################
def fit(self, X, y, learning_rate=0.2, epochs=10000):# x为训练集,y为标记,学习率,循环次数
X = np.atleast_2d(X) #二维
temp = np.ones([X.shape[0], X.shape[1]+1]) #设定形状
temp[:, 0:-1] = X # adding the bias unit to the input layer# 所有行,除去最后一列
X = temp #完成bias的添加
y = np.array(y) #数据类型转换
for k in range(epochs): #epochs次循环
i = np.random.randint(X.shape[0]) #每次随机抽取一行
a = [X[i]]
for l in range(len(self.weights)): #going forward network, for each layer
a.append(self.activation(np.dot(a[l], self.weights[l]))) #Computer the node value for each layer (O_i) using activation function
error = y[i] - a[-1] #计算误差
deltas = [error * self.activation_deriv(a[-1])] #For output layer, Err calculation (delta is updated error)
#Staring backprobagation 反向更新
for l in range(len(a) - 2, 0, -1): # 从最后一层开始,每次往回退一层
#Compute the updated error (i,e, deltas) for each node going from top layer to input layer
deltas.append(deltas[-1].dot(self.weights[l].T)*self.activation_deriv(a[l]))#更新隐藏层的函数
deltas.reverse()#逆向传播
for i in range(len(self.weights)):
layer = np.atleast_2d(a[i])
delta = np.atleast_2d(deltas[i])
self.weights[i] += learning_rate * layer.T.dot(delta) #更新权重的公式
#########################################################################################
def predict(self, x):
x = np.array(x)
temp = np.ones(x.shape[0]+1)
temp[0:-1] = x
a = temp
for l in range(0, len(self.weights)):
a = self.activation(np.dot(a, self.weights[l]))
return a
简单测试:
nn = NeuralNetwork( [2,2,1],'tanh' )
X = np.array( [[0,0],[0,1],[1,0],[1,1]] )
y = np.array( [0,1,1,0] )
nn.fit(X,y)
for i in [ [0,0],[0,1],[1,