<?xml version="1.0" encoding="utf-8" ?><rss version="2.0"><channel><title><![CDATA[evolution23的博客]]></title><description><![CDATA[]]></description><link>https://blog.youkuaiyun.com/evolution23</link><language>zh-cn</language><generator>https://blog.youkuaiyun.com/</generator><copyright><![CDATA[Copyright &copy; evolution23]]></copyright><item><title><![CDATA[ceph 数据清除]]></title><link>https://blog.youkuaiyun.com/evolution23/article/details/124129297</link><guid>https://blog.youkuaiyun.com/evolution23/article/details/124129297</guid><author>evolution23</author><pubDate>Tue, 12 Apr 2022 17:19:55 +0800</pubDate><description><![CDATA[lsblk -f

yum install gdisk

# Zap the disk to a fresh, usable state (zap-all is important, b/c MBR has to be clean)

# You will have to run this step for all disks.
sgdisk --zap-all /dev/sda

# Clean hdds with dd
dd if=/dev/zero of=/dev/sda bs=1M count=10]]></description><category></category></item><item><title><![CDATA[删除并清理rancher节点]]></title><link>https://blog.youkuaiyun.com/evolution23/article/details/124127320</link><guid>https://blog.youkuaiyun.com/evolution23/article/details/124127320</guid><author>evolution23</author><pubDate>Tue, 12 Apr 2022 16:23:40 +0800</pubDate><description><![CDATA[本文提供了一套完整的Docker和Kubernetes清理方案，主要包括：1.停止并删除所有Docker容器、卷；2.停止并禁用Kubernetes相关服务；3.卸载挂载目录并备份关键数据；4.删除残留路径和配置文件；5.清理网络接口和残留进程；6.重置iptables规则。该方案适用于需要彻底清理Docker和Kubernetes环境的场景，但部分命令（如删除所有镜像）需谨慎使用，且建议先备份重要数据。执行前应确认命令适用性，避免影响系统稳定性。]]></description><category></category></item><item><title><![CDATA[rancher +k8s 安装 rook ceph]]></title><link>https://blog.youkuaiyun.com/evolution23/article/details/118932877</link><guid>https://blog.youkuaiyun.com/evolution23/article/details/118932877</guid><author>evolution23</author><pubDate>Tue, 20 Jul 2021 13:37:20 +0800</pubDate><description><![CDATA[官网：https://www.rook.io/

前提条件：已部署完成 rancher+k8s系统，没有完成的请移步《Centos7 通过RKE 部署 高可用k8s+rancher》

一、镜像下载，我们需要将安装所需要的镜像从阿里云下载到本地（重要）

对应的版本号可查看 rook ceph中的operator.yaml配置文件





docker image pull rook/ceph:v1.6.7

docker pull registry.aliyuncs.com/it00021hot/ce.]]></description><category></category></item><item><title><![CDATA[Centos7 通过RKE 部署 高可用k8s+rancher]]></title><link>https://blog.youkuaiyun.com/evolution23/article/details/118928443</link><guid>https://blog.youkuaiyun.com/evolution23/article/details/118928443</guid><author>evolution23</author><pubDate>Tue, 20 Jul 2021 13:15:20 +0800</pubDate><description><![CDATA[前言：这里的rke rancher docker 是安装最新版的，如果需要安装特定版本，一些细节请自行百度和查询官网

环境介绍

操作系统：centos7

硬件配置：内存8GB以上 cpu2核以上 硬盘大于30G

系统架构：3个节点，每个节点同时承担master和work两种角色

10.9.1.200  rke-node1
10.9.1.201  rke-node2
10.9.1.201  rke-node3

操作步骤

（一）前期准备

注意以下操作需要在所有节点以root用...]]></description><category></category></item><item><title><![CDATA[Android上强大的图表库MPAndroidChart介绍]]></title><link>https://blog.youkuaiyun.com/evolution23/article/details/88387948</link><guid>https://blog.youkuaiyun.com/evolution23/article/details/88387948</guid><author>evolution23</author><pubDate>Mon, 11 Mar 2019 09:32:06 +0800</pubDate><description><![CDATA[转自《解析Android上强大的图表库MPAndroidChart》
]]></description><category></category></item><item><title><![CDATA[训练数据集与TFRecord互相转换]]></title><link>https://blog.youkuaiyun.com/evolution23/article/details/87855903</link><guid>https://blog.youkuaiyun.com/evolution23/article/details/87855903</guid><author>evolution23</author><pubDate>Thu, 21 Feb 2019 10:42:53 +0800</pubDate><description><![CDATA[转载自《训练数据集与TFRecord互相转换的两种方式》

TensorFlow使用TFRecord格式来统一存储数据，该格式可以将图像数据、标签信息、图像路径以及宽高等不同类型的信息放在一起进行统一存储，从而方便有效的管理不同的属性。

将训练数据集转成TFRecord

这里采用的数据集为目前正在做的项目的数据集，共包含两个目标文件夹(分别包含100幅图像)及对应的label.txt，labe...]]></description><category></category></item><item><title><![CDATA[机器学习总结——前言]]></title><link>https://blog.youkuaiyun.com/evolution23/article/details/87274560</link><guid>https://blog.youkuaiyun.com/evolution23/article/details/87274560</guid><author>evolution23</author><pubDate>Thu, 14 Feb 2019 16:19:24 +0800</pubDate><description><![CDATA[机器学习任务类型主要分为两大类：
        • 监督学习(Supervised Learning)
        – 分类 (Classification)
        – 回归 (Regression)
        – 排序 (Ranking)
        • 非监督学习 (Unsupervised Learning)
        – 聚类 (Clustering)
  ...]]></description><category></category></item><item><title><![CDATA[Tensorflow入门教程]]></title><link>https://blog.youkuaiyun.com/evolution23/article/details/86408188</link><guid>https://blog.youkuaiyun.com/evolution23/article/details/86408188</guid><author>evolution23</author><pubDate>Sun, 13 Jan 2019 10:09:26 +0800</pubDate><description><![CDATA[《比官方更简洁的Tensorflow入门教程》]]></description><category></category></item><item><title><![CDATA[sklearn 中的MiniBatchKMeans(聚类)使用]]></title><link>https://blog.youkuaiyun.com/evolution23/article/details/85449887</link><guid>https://blog.youkuaiyun.com/evolution23/article/details/85449887</guid><author>evolution23</author><pubDate>Mon, 31 Dec 2018 10:18:26 +0800</pubDate><description><![CDATA[1、前期准备


#导入必要的工具包
import pandas as pd
import numpy as np
from sklearn.cluster import MiniBatchKMeans
from sklearn.model_selection import train_test_split
from sklearn import metrics

from sklearn.dec...]]></description><category></category></item><item><title><![CDATA[sklearn的PCA使用]]></title><link>https://blog.youkuaiyun.com/evolution23/article/details/85448613</link><guid>https://blog.youkuaiyun.com/evolution23/article/details/85448613</guid><author>evolution23</author><pubDate>Mon, 31 Dec 2018 10:09:35 +0800</pubDate><description><![CDATA[1、数据准备


#导入必要的工具包
import pandas as pd
import numpy as np
from sklearn import svm
from sklearn.model_selection import train_test_split
from sklearn.decomposition import PCA
import time

#读取训练数据和测试数据
t...]]></description><category></category></item><item><title><![CDATA[LightGBM使用指南]]></title><link>https://blog.youkuaiyun.com/evolution23/article/details/85375310</link><guid>https://blog.youkuaiyun.com/evolution23/article/details/85375310</guid><author>evolution23</author><pubDate>Sun, 30 Dec 2018 11:18:35 +0800</pubDate><description><![CDATA[1、准备


import lightgbm as lgbm
import pandas as pd 
import numpy as np
from sklearn.model_selection import GridSearchCV
from sklearn.metrics import log_loss
from matplotlib import pyplot
import seabor...]]></description><category></category></item><item><title><![CDATA[Pandas速查手册]]></title><link>https://blog.youkuaiyun.com/evolution23/article/details/85284735</link><guid>https://blog.youkuaiyun.com/evolution23/article/details/85284735</guid><author>evolution23</author><pubDate>Thu, 27 Dec 2018 15:22:38 +0800</pubDate><description><![CDATA[本文转自《Pandas速查手册中文版》

对于数据科学家，无论是数据分析还是数据挖掘来说，Pandas是一个非常重要的Python包。它不仅提供了很多方法，使得数据处理非常简单，同时在数据处理速度上也做了很多优化，使得和Python内置方法相比时有了很大的优势。

如果你想学习Pandas，建议先看两个网站。

（1）官网：Python Data Analysis Library

（2）十分钟入...]]></description><category></category></item><item><title><![CDATA[决策树参数调优建议]]></title><link>https://blog.youkuaiyun.com/evolution23/article/details/85253992</link><guid>https://blog.youkuaiyun.com/evolution23/article/details/85253992</guid><author>evolution23</author><pubDate>Tue, 25 Dec 2018 20:57:27 +0800</pubDate><description><![CDATA[建议一：

三个最重要的参数为:树的数目、树的深度和学习率。建议参数调整策略为:
– 采用默认参数配置试试
– 如果系统过拟合了,降低学习率
– 如果系统欠拟合,加大学习率

建议二：

– n_estimators和learning_rate:固定n_estimators为100(数目不大，因为树的深度较大,每棵树比较复杂),然后调整learning_rate
– 树的深度max_depth:从...]]></description><category></category></item><item><title><![CDATA[sklearn中XGBoost的使用]]></title><link>https://blog.youkuaiyun.com/evolution23/article/details/85253608</link><guid>https://blog.youkuaiyun.com/evolution23/article/details/85253608</guid><author>evolution23</author><pubDate>Tue, 25 Dec 2018 20:43:19 +0800</pubDate><description><![CDATA[1、数据准备


from xgboost import XGBClassifier
import xgboost as xgb

import pandas as pd 
import numpy as np

from sklearn.model_selection import GridSearchCV
from sklearn.model_selection import Stratifi...]]></description><category></category></item><item><title><![CDATA[sklearn中随机森林的使用]]></title><link>https://blog.youkuaiyun.com/evolution23/article/details/85244195</link><guid>https://blog.youkuaiyun.com/evolution23/article/details/85244195</guid><author>evolution23</author><pubDate>Tue, 25 Dec 2018 10:39:33 +0800</pubDate><description><![CDATA[1、数据准备与《sklearn中决策树的使用》中相同，这里不再累述、

2、使用步骤


from sklearn.ensemble import RandomForestClassifier
model_RR=RandomForestClassifier()
model_RR.fit(X_train,y_train)

y_prob = model_RR.predict_proba(X_test...]]></description><category></category></item><item><title><![CDATA[sklearn中决策树的GridSearchCV参数调优]]></title><link>https://blog.youkuaiyun.com/evolution23/article/details/85243980</link><guid>https://blog.youkuaiyun.com/evolution23/article/details/85243980</guid><author>evolution23</author><pubDate>Tue, 25 Dec 2018 10:34:12 +0800</pubDate><description><![CDATA[决策树的超参数有：

max_depth（树的深度）
	max_leaf_nodes（叶子结点的数目）
	max_features（最大特征数目）
	min_samples_leaf（叶子结点的最小样本数）
	min_samples_split（中间结点的最小样本树）
	min_weight_fraction_leaf（叶子节点的样本权重占总权重的比例）
	min_impurity_split（最...]]></description><category></category></item><item><title><![CDATA[sklearn中决策树的使用]]></title><link>https://blog.youkuaiyun.com/evolution23/article/details/85235397</link><guid>https://blog.youkuaiyun.com/evolution23/article/details/85235397</guid><author>evolution23</author><pubDate>Tue, 25 Dec 2018 09:54:28 +0800</pubDate><description><![CDATA[1、数据准备


import pandas as pd
import numpy as np
from matplotlib import pyplot as plt
from sklearn.model_selection import GridSearchCV
from sklearn.metrics import accuracy_score
from sklearn.metrics im...]]></description><category></category></item><item><title><![CDATA[人工智能—XGBoost使用指南]]></title><link>https://blog.youkuaiyun.com/evolution23/article/details/85196173</link><guid>https://blog.youkuaiyun.com/evolution23/article/details/85196173</guid><author>evolution23</author><pubDate>Sat, 22 Dec 2018 10:06:50 +0800</pubDate><description><![CDATA[步骤：

            1. 读取数据--&amp;gt; DMatrix
            2. 设置参数
            3. 模型训练：train/cv
                3.1: train with 在校验集上early stop
                3.2: cv
            4. 预测   

1、读取数据

XGBoost可以加...]]></description><category></category></item><item><title><![CDATA[Logistic 回归—SVM正则参数调优操作笔记]]></title><link>https://blog.youkuaiyun.com/evolution23/article/details/85045740</link><guid>https://blog.youkuaiyun.com/evolution23/article/details/85045740</guid><author>evolution23</author><pubDate>Mon, 17 Dec 2018 10:56:59 +0800</pubDate><description><![CDATA[1、准备


# 首先 import 必要的模块
import pandas as pd 
import numpy as np

from sklearn.model_selection import GridSearchCV

#竞赛的评价指标为logloss
#from sklearn.metrics import log_loss  
#SVM并不能直接输出各类的概率，所以在这个例子中我们...]]></description><category></category></item><item><title><![CDATA[Logistic 回归—LogisticRegressionCV实现参数优化]]></title><link>https://blog.youkuaiyun.com/evolution23/article/details/85028423</link><guid>https://blog.youkuaiyun.com/evolution23/article/details/85028423</guid><author>evolution23</author><pubDate>Sun, 16 Dec 2018 10:02:05 +0800</pubDate><description><![CDATA[1、准备


# 首先 import 必要的模块
import pandas as pd 
import numpy as np
 
from sklearn.model_selection import GridSearchCV
 
#竞赛的评价指标为logloss
from sklearn.metrics import log_loss  
 
from matplotlib import p...]]></description><category></category></item></channel></rss>