▍1. Kares入门资料打包
深度学习框架Keras入门资料,里面的代码包括课件和DEMO有利于新书入门学习,简单易懂(Keras Introductory Information of Deep Learning Framework, which includes courseware and DEMO, is helpful for introductory learning of new books. It is easy to understand.)
深度学习框架Keras入门资料,里面的代码包括课件和DEMO有利于新书入门学习,简单易懂(Keras Introductory Information of Deep Learning Framework, which includes courseware and DEMO, is helpful for introductory learning of new books. It is easy to understand.)
说明: 深度学习框架Keras入门资料,里面的代码包括课件和DEMO有利于新书入门学习,简单易懂(Keras Introductory Information of Deep Learning Framework, which includes courseware and DEMO, is helpful for introductory learning of new books. It is easy to understand.)
boxcox函数的python实现,引用该函数可将偏态分布调整为正态分布(Python implementation of box Cox function)
说明: boxcox函数的python实现,引用该函数可将偏态分布调整为正态分布(Python implementation of box Cox function)
利用python进行数据分析,英文书籍,从pandas库的数据分析工具开始利用高性能工具对数据进行加载、清理、转换、合并以及重塑;利用matpIotlib创建散点图以及静态或交互式的可视化结果;利用pandas的groupby功能对数据集进行切片、切块和汇总操作;处理各种各样的时间序列数据。(Starting from the data analysis tools of pandas database, high performance tools are used to load, clean, transform, merge and remodel data; scatter plots and static or interactive visualization results are created by matpIotlib; data sets are sliced, sliced and aggregated by pandas group by function; and various operations are processed. Time series data.)
说明: 利用python进行数据分析,英文书籍,从pandas库的数据分析工具开始利用高性能工具对数据进行加载、清理、转换、合并以及重塑;利用matpIotlib创建散点图以及静态或交互式的可视化结果;利用pandas的groupby功能对数据集进行切片、切块和汇总操作;处理各种各样的时间序列数据。(Starting from the data analysis tools of pandas database, high performance tools are used to load, clean, transform, merge and remodel data; scatter plots and static or interactive visualization results are created by matpIotlib; data sets are sliced, sliced and aggregated by pandas group by function; and various operations are processed. Time series data.)
说明: 用于轴承大数据的故障诊断和数据挖掘,可将轴承的振动信息进行数组分析,获得预测模型,准确率较高(It can be used for fault diagnosis and data mining of bearing big data. It can analyze the vibration information of bearing by array and obtain the prediction model with high accuracy)
说明: 用python 做的网络数据爬虫,爬取淘宝数据,并分析。(Use Python to do the network data crawler, crawl the Taobao data, and analyze.)
说明: 一种无监督的聚类算法,基于密度聚类,名称为基于快速搜索与寻找密度峰值的聚类(Clustering by fast search and find of desity peaks)
说明: 采用机器学习预测房价.使用波士顿房屋信息数据来训练和测试一个模型,并对模型的性能和预测能力进行评估。(Using Machine Learning to Predict House Prices)
隐马尔科夫实现,包含forward-hmm, Viterbi-hmm, Baum-Welch-hmm(Hidden Markov implementation, including forward-hmm, Viterbi-hmm, Baum-Welch-hmm)
说明: 隐马尔科夫实现,包含forward-hmm, Viterbi-hmm, Baum-Welch-hmm(Hidden Markov implementation, including forward-hmm, Viterbi-hmm, Baum-Welch-hmm)
《从零开始学爬虫》的配套资料(PPT和源码)("Learning Reptiles from Zero" (PPT and Source))
说明: 《从零开始学爬虫》的配套资料(PPT和源码)("Learning Reptiles from Zero" (PPT and Source))
很好用的关联规则挖掘经典算法,推荐使用。包括支持度、置信度、提升度,输出结果到excel文件(Good use of association rules mining classic algorithm, recommended)
kNN算法的核心思想是如果一个样本在特征空间中的k个最相邻的样本中的大多数属于某一个类别,则该样本也属于这个类别,并具有这个类别上样本的特性。该方法在确定分类决策上只依据最邻近的一个或者几个样本的类别来决定待分样本所属的类别。 kNN方法在类别决策时,只与极少量的相邻样本有关。由于kNN方法主要靠周围有限的邻近的样本,而不是靠判别类域的方法来确定所属类别的,因此对于类域的交叉或重叠较多的待分样本集来说,kNN方法较其他方法更为适合。(Basic source application)
机器学习 数据挖掘 数值算法 人工智能 全英文教材(python machine learning data snooping)
决策树与集成算法,用来分类已知数据种类,希望对编程有帮助(Decision tree and ensemble algorithm are used to classify known data types.)