rmx140jvmc9出现22-tpp是什么意思思

来自CSDN博客:针对TAB切换的隐藏区域
blog__166300
不能显示此文件
来自CSDN博客:Pearson product-moment correlation coefficient in java(java的简单相关系数算法)
blog__4839672
*/package numerator.pearson.conefficient;import java.util.ArrayList;import java.util.List;import java.io.BufferedReader;import java.io.IOException;import java.io.InputStreamReader;/** * @author alan-king *
* the class is going to cal *
* */public class NumeratorCalculate {
//add global varieties protected List&String& xList , yList;
public NumeratorCalculate(List&String& xList ,List&String& yList){
this.xList = xList;
this.yList = yList; }
* add operate method
*/ public double calcuteNumerator(){
double result =0.0;
double xAverage = 0.0;
double temp = 0.0;
int xSize = xList.size();
for(int x=0;x&xSize;x++){
temp += Double.parseDouble(xList.get(x));
xAverage = temp/xSize;
double yAverage = 0.0;
temp = 0.0;
int ySize = yList.size();
for(int x=0;x&ySize;x++){
temp += Double.parseDouble(yList.get(x));
yAverage = temp/ySize;
//double sum = 0.0;
for(int x=0;x&xSize;x++){
result+=(Double.parseDouble(xList.get(x))-xAverage)*(Double.parseDouble(yList.get(x))-yAverage);
return result; }}
来自CSDN博客:bagging算法java实现(从N个样本中有放回地取N次)
blog__4293142
package cn.melina.classification.test;import java.text.DecimalFormat;import java.util.ArrayList;import java.util.HashMap;import java.util.Random;import java.util.Set;import org.mon.RandomUtils;public class BaggingTest { /**
* if data has N cases, sample N cases at random without replacement.
* @author melina
* @param N
numbers of cases
* @return N次取值并去重之后剩余的数的个数
*/ public static int runBagging(int N){
Random rng = RandomUtils.getRandom();
ArrayList&Integer& list = new ArrayList&Integer&();
for (int i = 0; i & N; i++) {
int index = rng.nextInt(N);
list.add(index);
//合并相同的取值
HashMap&String, Integer& hash = new HashMap&String, Integer&();
for (int i = 0; i & list.size(); i++) {
if (!hash.isEmpty() && hash.containsKey(list.get(i))) {
hash.put(list.get(i).toString(), hash.get(list.get(i)) + 1);
hash.put(list.get(i).toString(), 1);
} catch (Exception e) {
/*Set&String& set = hash.keySet();
for (String key : set) {
System.out.println(key + &==&& + hash.get(key));
return hash.keySet().size(); }
public static void main(String []args){
int itr_num = 10000;
//迭代次数
int datasize = 100;
//bagging的样本数目,此处为0~99之间100个数字做bagging
ArrayList&Integer& list = new ArrayList&Integer&();
for(int i = 0; i & itr_num; i ++){
int num = runBagging(datasize);
list.add(num);
//System.out.println(&第&+i+&次bagging去重之后的个数:&+ num);
//System.out.println(num);
//统计 相同的数目在全部迭代后出现的频率
HashMap&String, Integer& hash = new HashMap&String, Integer&();
for (int i = 0; i & list.size(); i++) {
if ((!hash.isEmpty() )&&( hash.containsKey(list.get(i).toString()))) {
hash.put(list.get(i).toString(), Integer.valueOf(hash.get(list.get(i).toString())) + 1);
hash.put(list.get(i).toString(), 1);
} catch (Exception e) {
Set&String& set = hash.keySet();
for (String key : set) {
double itr_double=itr_num*1.0;
double value =
hash.get(key)/itr_double;
DecimalFormat df = new DecimalFormat(&0.00%&);
System.out.println(key + &,& + df.format(value));
来自CSDN博客:Jackknife 刀切法
blog__8445676
function out=jackbias(theta,orig)%Estimate the bias using the jackknife%Theta has to be a character string containg% a valid function name[n,p]=size(orig);lot=feval(theta,orig(2:n,:));k=length(lot);lo=zeros(n,k);lo(1,:)=lo(n,:)=feval(theta,orig(1:(n-1),:));for i=(2:(n-1))
lo(i,:)=feval(theta,orig([1:(i-1),(i+1):n],:)); endthetadot=mean(lo);out=(n-1)*(thetadot-feval(theta,orig));
来自CSDN博客:scikit-learn: Machine Learning in Python——Supervised learning[一](3)
blog__7260362
In [1]: from sklearn import linear_modelIn [2]: clf = linear_model.Rlinear_model.RandomizedLassolinear_model.RandomizedLogisticRegressionlinear_model.Ridgelinear_model.RidgeCVlinear_model.RidgeClassifierlinear_model.RidgeClassifierCVIn [2]: clf = linear_model.Ridge(alpha = .5)In [3]: clf.fit([[0, 0], [0, 0], [1, 1]], [0, .1, 1])Out[3]: Ridge(alpha=0.5, copy_X=True, fit_intercept=True, max_iter=None,
normalize=False, solver='auto', tol=0.001)In [4]: clf.coef_Out[4]: array([ 0.,
0.])In [5]: clf.intercept_Out[5]: 0.63641
来自CSDN博客:《统计学习方法——李航》学习小记——perceptron
blog__6836623
&span style=&font-size:18&&
&/span&&/p&&p&&span style=&font-size:18&&
感知机,作为后面支持向量机等的学习基础,很有必要好好学学研究透。只能用来在特征空间中线性分类,属于判别模型,形式为:&/span&&div style=&text-align:&&w·x+b=0&/div&
这里w·x为两个向量的点乘,还可以写成w*x^T这里是乘上x的转置。对于一个线性可分的数据集,这里,李航的书为了简便并未用粗体来区分向量和标量,而国外的教程中一般通过粗体来表示出向量,正常字体表示标量。
感知机的学习策略就是通过从一个初始的超平面开始,通过错分类点,逐步逐步调整参数w和b来使错分类点到达超平面的正确的一边。这里要注意一下一些方面:(1)随机梯度下降和梯度下降的区别
调整策略使用的是随机梯度下降法(stochastic gradient descent),它和梯度下降的区别如下Gradient Descent: You need to run over every training example before doing an update, which means that if you have a large dataset, you might spend much time on getting something that works.Stochastic gradient descent:on the other hand, does updates every time it finds a training example, however, since it only uses one update, it may never converge, although you can still be pretty close to the minimum.简言之就是梯度下降是把所有样本带入计算,而随机梯度下降每次只使用一个样本,当样本很大时迭代一次的速度远远大于梯度下降。随机梯度下降:&img src=&data:image/base64,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& alt=&& /&梯度下降:&img src=&data:image/base64,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& alt=&& /&(2)问题的原始形式和对偶形式
每一个线性规划问题都伴随着另一个线性规划问题,我们称之为对偶问题,原问题就称为原始问题。对偶的基本想法是将原始问题中的参数表示为样本实例和标记的线性组合形式,通过求解系数来求解原参数,这里有句话叫做实例点更新次数越多,意味着它离超平面越近,即越难分类。
贴上书上例题对偶形式的算法:正样本点(3,3),(4,3),负样本点(1,1)&/p&&p&&span style=&font-size: 18&&&/span&&/p&&p&&span style=&font-size:18&&&/span&&/p&&p style=&text-align:left&&&span style=&font-size:18px&&&/span&&/p&&p style=&text-align:left&&&span style=&font-size:18px&&&/span&&/p&&pre code_snippet_id=&625128& snippet_file_name=&blog__1980285& name=&code& class=&cpp&&#include&iostream&using namespace std;int main(){ int x[3][2]={
{1,1} }; int y[3]={1,1,-1}; int alp[3]={0}; int g[3][3],k,m,b,flag; k=b=0; flag=1;
for(int i=0;i&3;i++)
for(int j=0;j&3;j++)
g[i][j]=x[i][0]*x[j][0]+x[i][1]*x[j][1];
while(flag)
for(int i=0;i&3;i++)
while(true)
for(int j=0;j&3;j++)
m+=alp[j]*y[j]*g[j][i];
for(int i=0;i&3;i++)
cout&&&alpa&&&i+1&&&: &&&alp[i]&&endl; cout&&&b: &&&b&&endl;}
来自CSDN博客:scikit-learn: Machine Learning in Python——Supervised learning(一)(3)
blog__6268387
In [1]: from sklearn import linear_modelIn [2]: clf = linear_model.Rlinear_model.RandomizedLassolinear_model.RandomizedLogisticRegressionlinear_model.Ridgelinear_model.RidgeCVlinear_model.RidgeClassifierlinear_model.RidgeClassifierCVIn [2]: clf = linear_model.Ridge(alpha = .5)In [3]: clf.fit([[0, 0], [0, 0], [1, 1]], [0, .1, 1])Out[3]: Ridge(alpha=0.5, copy_X=True, fit_intercept=True, max_iter=None,
normalize=False, solver='auto', tol=0.001)In [4]: clf.coef_Out[4]: array([ 0.,
0.])In [5]: clf.intercept_Out[5]: 0.63641
来自CSDN博客:《算法导论》10、中位数和顺序统计学(C++)
blog__5476969
#include &iostream&
#include &stdlib.h&
#include&time.h&
using namespace std;int partition(int* A, int p, int r){ if (p & r) {
int r1 = (rand() % (r - p + 1)) + p;
int temp = A[r1];
A[r1] = A[r];
A[r] = temp;
int x = A[r];
int i = p - 1, j;
for (j = p; j & r; j++)
if (A[j] &= x)
temp = A[i];
A[i] = A[j];
A[j] = temp;
temp = A[i + 1];
A[i + 1] = A[r];
A[r] = temp;
return i + 1; }}int quickSelect(int* A, int p, int r, int i){ if (p == r)
return A[p]; int q = partition(A, p, r); int k = q - p + 1; if (i == k)
return A[q]; else if (i & k)
return quickSelect(A, p, q - 1, i); else
return quickSelect(A, q + 1, r, i - k);}void main(){ int n = 100;
//排序元素长度
int* A = new int[n]; for (int i = 0; i &n; i++) {
A[i] = rand() % 1000;
cout && A[i] && &
&; } cout && endl; clock_t start, finish; double totaltime; start = clock(); cout && quickSelect(A, 0, n - 1, 1) && endl;
for (int i = 0; i &n; i++) {
cout && A[i] && &
&; } cout && endl; finish = clock(); totaltime = (double)(finish - start) / CLOCKS_PER_SEC; cout && &运行时间为& && totaltime && &秒!& && endl; system(&pause&);}}

我要回帖

更多关于 是什么意思 的文章

更多推荐

版权声明:文章内容来源于网络,版权归原作者所有,如有侵权请点击这里与我们联系,我们将及时删除。

点击添加站长微信