liuguaaa的博客

原创 k-means聚类算法

clear all;clc;tic; X=[ 91.8244186,22.3705287; 92.0069949,21.4282895; 90.3917291,23.764319; 90.3858206,22.4929908; 91.8780611,24.9015238; 88.6074625,24.3750433; 89.2166017,23.1468977; ]; k=2;[Idx,Ctrs,SumD,D]=kmeans(X,k); plot(X(Idx=

原创 随机梯度下降法

import numpy as npimport matplotlib.pyplot as pltfrom mpl_toolkits.mplot3d import Axes3De0 = 1e1 = 1e2 = 1alpha = 0.01theta0 = np.random.uniform(0, 1)theta1 = np.random.uniform(0, 1)theta2 = np.random.uniform(0, 1)theta = np.array([theta1, thet

原创 决策树中信息熵和信息增益的推导

原创 Actor Critic Method算法(强化学习)

import gym,osfrom itertools import countimport paddleimport paddle.nn as nnimport paddle.optimizer as optimimport paddle.nn.functional as Ffrom paddle.distribution import Categoricaldevice = paddle.get_device()env = gym.make("CartPole-v0")stat

原创 主成分分析PCA(降维)

import numpy as np#初始化矩阵data = np.array([ [2.5,2.4], [0.5,0.7], [2.2,2.9], [1.9,2.2], [3.1,3.0], [2.3,2.7], [2, 1.6],

原创 全连接神经网络

import numpy as npimport randomimport sysimport mathalpha = 0.01#sigmoid函数def sigmoid(x): return 1 / (1 + math.exp(-x))#损失函数def loss(): sigema1 = (y[0] - t[0]) ** 2 sigema2 = (y[1] - t[1]) ** 2 return sigema1 + sigema2#t是预测值

原创 改进欧拉算法

import matplotlib.pyplot as plty = 1x = 0xx = [0] * 101yy = [0] * 101k1 = y - 0.2 * x / y #k1存储当前点的斜率k2 = 0.00 #k2存储下一个点的斜率for i in range (1,101): if(i == 1): #第一次计算的时候只有k1所以需要用朴素欧拉计算第一次的值 y = 1.1 * y - 0.2 *

原创 全连接神经网络公式推导

原创 用segmoid函数拟合阶跃函数

import matplotlib.pyplot as pltfrom mpl_toolkits.mplot3d import Axes3Dimport mathimport randomdef sigmoid(x): return 1 / (1 + math.exp(-x))e0 = 1e1 = 1e2 = 1alpha = 0.001omega0 = np.random.uniform(0,1)omega1 = np.random.uniform(0,1)omeg

翻译 numpy操纵代码的还原

c = np.arange(1,13).reshape(6,2)carray([[ 1, 2], [ 3, 4], [ 5, 6], [ 7, 8], [ 9, 10], [11, 12]])np.vsplit(c,3)[array([[1, 2], [3, 4]]), array([[5, 6], [7, 8]]), array([[ 9, 10], [1...

原创 机器学习：用matplotlib绘制拟合曲线

import numpy as npimport matplotlib.pyplot as pltfrom mpl_toolkits.mplot3d import Axes3De0 = 1e1 = 1e2 = 1alpha = 0.01theta0 = np.random.uniform(0,1)theta1 = np.random.uniform(0,1)theta2 = np.random.uniform(0,1)theta = np.array([theta1, thet.