吕雄

2018

March

书籍不如钱一囊，少年何苦擅文章

领解皇都第一名，猖披归卧旧茅蘅。

立锥莫笑无余地，万里江山笔下生。

                  ——唐寅《无题》

诗文圣手，书画双绝。他风流不羁才气无双，他也曾失意长叹举杯惆怅;他飘逸潇洒纵酒狂歌，他也曾历经沧桑郁郁寡欢。三分豪迈七分悲凉，这一生，他只有一个心愿：“但愿老死花酒间，不愿鞠躬车马前”。这一世，他只做一件事：“桃花仙人种桃树，又摘桃花卖酒钱。”

March 5 , 2018 樵林后子阅读全文

3、训练一个神经网络 搭建由一个输入层，一个隐藏层，一个输出层组成的三层神经网络。输入层中的节点数由数据的维度来决定，也就是2个。相应的，输出层的节点数则是由类的数量来决定，也是2个（因为我们只有一个预测0和1的输出节点，所以我们只有两类输出，实际中，两个输出节点将更易于在后期进行扩展从而获得更多类别的输出）以X，Y坐标作为输入，输出的则是两种概率，一种是0（代表女），另一种是1（代表男）结果如下。：

搭建神经网络

num_examples = len(X) # training set size
nn_input_dim = 2 # input layer dimensionality
nn_output_dim = 2 # output layer dimensionality

# Gradient descent parameters (I picked these by hand)
epsilon = 0.01 # learning rate for gradient descent
reg_lambda = 0.01 # regularization strength
# Helper function to evaluate the total loss on the dataset
def calculate_loss(model):
    W1, b1, W2, b2 = model['W1'], model['b1'], model['W2'], model['b2']
    # Forward propagation to calculate our predictions
    z1 = X.dot(W1) + b1
    a1 = np.tanh(z1)
    z2 = a1.dot(W2) + b2
    exp_scores = np.exp(z2)
    probs = exp_scores / np.sum(exp_scores, axis=1, keepdims=True)
    # Calculating the loss
    corect_logprobs = -np.log(probs[range(num_examples), y])
    data_loss = np.sum(corect_logprobs)
    # Add regulatization term to loss (optional)
    data_loss += reg_lambda/2 * (np.sum(np.square(W1)) + np.sum(np.square(W2)))
    return 1./num_examples * data_loss
# Helper function to predict an output (0 or 1)
def predict(model, x):
    W1, b1, W2, b2 = model['W1'], model['b1'], model['W2'], model['b2']
    # Forward propagation
    z1 = x.dot(W1) + b1
    a1 = np.tanh(z1)
    z2 = a1.dot(W2) + b2
    exp_scores = np.exp(z2)
    probs = exp_scores / np.sum(exp_scores, axis=1, keepdims=True)
    return np.argmax(probs, axis=1)
# This function learns parameters for the neural network and returns the model.
# - nn_hdim: Number of nodes in the hidden layer
# - num_passes: Number of passes through the training data for gradient descent
# - print_loss: If True, print the loss every 1000 iterations
def build_model(nn_hdim, num_passes=20000, print_loss=False):

    # Initialize the parameters to random values. We need to learn these.
    np.random.seed(0)
    W1 = np.random.randn(nn_input_dim, nn_hdim) / np.sqrt(nn_input_dim)
    b1 = np.zeros((1, nn_hdim))
    W2 = np.random.randn(nn_hdim, nn_output_dim) / np.sqrt(nn_hdim)
    b2 = np.zeros((1, nn_output_dim))

    # This is what we return at the end
    model = {}

    # Gradient descent. For each batch...
    for i in range(0, num_passes):

        # Forward propagation
        z1 = X.dot(W1) + b1
        a1 = np.tanh(z1)
        z2 = a1.dot(W2) + b2
        exp_scores = np.exp(z2)
        probs = exp_scores / np.sum(exp_scores, axis=1, keepdims=True)

        # Backpropagation
        delta3 = probs
        delta3[range(num_examples), y] -= 1
        dW2 = (a1.T).dot(delta3)
        db2 = np.sum(delta3, axis=0, keepdims=True)
        delta2 = delta3.dot(W2.T) * (1 - np.power(a1, 2))
        dW1 = np.dot(X.T, delta2)
        db1 = np.sum(delta2, axis=0)

        # Add regularization terms (b1 and b2 don't have regularization terms)
        dW2 += reg_lambda * W2
        dW1 += reg_lambda * W1

        # Gradient descent parameter update
        W1 += -epsilon * dW1
        b1 += -epsilon * db1
        W2 += -epsilon * dW2
        b2 += -epsilon * db2

        # Assign new parameters to the model
        model = { 'W1': W1, 'b1': b1, 'W2': W2, 'b2': b2}

        # Optionally print the loss.
        # This is expensive because it uses the whole dataset, so we don't want to do it too often.
        if print_loss and i % 1000 == 0:
          print ("Loss after iteration %i: %f" %(i, calculate_loss(model)))

    return model
# Build a model with a 3-dimensional hidden layer
model = build_model(3, print_loss=True)

# Plot the decision boundary
plot_decision_boundary(lambda x: predict(model, x))
plt.title("Decision Boundary for hidden layer size 3")

隐藏层规模为3的网络表

隐藏层规模为3的网络图

4、变更隐藏层规模(5图)

plt.figure(figsize=(16, 32))
hidden_layer_dimensions = [1, 2, 3, 4, 5, 20, 50]
for i, nn_hdim in enumerate(hidden_layer_dimensions):
    plt.subplot(5, 2, i+1)
    plt.title('Hidden Layer size %d' % nn_hdim)
    model = build_model(nn_hdim)
    plot_decision_boundary(lambda x: predict(model, x))
plt.show()

变更隐藏层规模

March 1 , 2018 阅读全文

February

用python自动刷票

# -*- coding: utf-8 -*-
"""
Created on Fri Feb 23 20:19:41 2018

@author: lenovo
"""
import re
import random
import sys
import time
import datetime
import threading
from random import choice
import requests
import bs4


def get_ip():
    """获取代理IP"""
    url = "http://www.xicidaili.com/nn"
    headers = { "Accept":"text/html,application/xhtml+xml,application/xml;",
                "Accept-Encoding":"gzip, deflate, sdch",
                "Accept-Language":"zh-CN,zh;q=0.8,en;q=0.6",
                "Referer":"http://www.xicidaili.com",
                "User-Agent":"Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/42.0.2311.90 Safari/537.36"
                }
    r = requests.get(url,headers=headers)
    soup = bs4.BeautifulSoup(r.text, 'html.parser')
    data = soup.table.find_all("td")
    ip_compile= re.compile(r'<td>(\d+\.\d+\.\d+\.\d+)</td>')    # 匹配IP
    port_compile = re.compile(r'<td>(\d+)</td>')                # 匹配端口
    ip = re.findall(ip_compile,str(data))       # 获取所有IP
    port = re.findall(port_compile,str(data))   # 获取所有端口
    return [":".join(i) for i in zip(ip,port)]  # 组合IP+端口，如：115.112.88.23:8080


# 设置 user-agent列表，每次请求时，可在此列表中随机挑选一个user-agnet
uas = [
    "Mozilla/5.0 (X11; Ubuntu; Linux x86_64; rv:17.0; Baiduspider-ads) Gecko/17.0 Firefox/17.0",
    "Mozilla/5.0 (Windows; U; Windows NT 5.1; zh-CN; rv:1.9b4) Gecko/2008030317 Firefox/3.0b4",
    "Mozilla/5.0 (Windows; U; MSIE 6.0; Windows NT 5.1; SV1; .NET CLR 2.0.50727; BIDUBrowser 7.6)",
    "Mozilla/5.0 (Windows NT 6.3; WOW64; Trident/7.0; rv:11.0) like Gecko",
    "Mozilla/5.0 (Windows NT 6.3; WOW64; rv:46.0) Gecko/20100101 Firefox/46.0",
    "Mozilla/5.0 (Windows NT 6.3; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/45.0.2454.99 Safari/537.36",
    "Mozilla/5.0 (Windows NT 6.3; Win64; x64; Trident/7.0; Touch; LCJB; rv:11.0) like Gecko",
    ]


def get_url(code=0,ips=[]):
    """
        投票
        如果因为代理IP不可用造成投票失败，则会自动换一个代理IP后继续投
    """
    try:
        ip = choice(ips)
    except:
        return False
    else:
        proxies = {
            "http":ip,
        }
        headers2 = {
            "Accept":"*/*",
            "Accept-Encoding":"gzip, deflate, sdch",
            "Accept-Language":"zh-CN,zh;q=0.8,en;q=0.6",
            "Referer":"https://best.zhaopin.com/",
            "User-Agent":choice(uas),
        }
        datas = {'bestid': 11174, 'source': 'best'}
    try:
        hz_url = "https://best.zhaopin.com/API/Vote.ashx"   # 某投票网站的地址，这里不用真实的域名
        hz_r = requests.post(hz_url,headers=headers2,data=datas,proxies=proxies)
    except requests.exceptions.ConnectionError:
        print ('ConnectionError')
        if not ips:
            print ('not ip')
            sys.exit()
        # 删除不可用的代理IP
        if ip in ips:
            ips.remove(ip)
        # 重新请求URL
        get_url(code,ips)
    else:
        date = datetime.datetime.now().strftime('%H:%M:%S')
        print ("code={0},date={1},ip={2},hz_r.text={3},len(ips)={4}".format(code,date,ip,hz_r.text,len(ips)))


ips = []
for i in range(6000):
    # 每隔1000次重新获取一次最新的代理IP，每次可获取最新的100个代理IP
    if i % 1000 == 0:
        ips.extend(get_ip())
    # 启用线程，隔1秒产生一个线程，可控制时间加快投票速度 ,time.sleep的最小单位是毫秒
    t1 = threading.Thread(target=get_url,args=(i,ips))
    t1.start()
    time.sleep(1)

原文链接：https://www.cnblogs.com/zhouxinfei/p/7861966.html

February 23 , 2018 阅读全文