摘要:
In this post we take a tour of the most popular machine learning algorithms. It is useful to tour the main algorithms in the field to get a feeling of 阅读全文
随笔分类 - Machine Learning
Maximum Likelihood Method最大似然法
2016-03-01 16:34 by GarfieldEr007, 567 阅读, 收藏,
摘要:
最大似然法,英文名称是Maximum Likelihood Method,在统计中应用很广。这个方法的思想最早由高斯提出来,后来由菲舍加以推广并命名。 最大似然法是要解决这样一个问题:给定一组数据和一个参数待定的模型,如何确定模型的参数,使得这个确定参数后的模型在所有模型中产生已知数据的概率最 大。 阅读全文
最小二乘法least square
2016-03-01 11:37 by GarfieldEr007, 628 阅读, 收藏,
摘要:
上研究生的时候接触的第一个Loss function就是least square。最近又研究了一下,做个总结吧。 定义看wiki就够了。公式如下 E(w)=12∑n=1N{y−xWT}2E(w)=12∑n=1N{y−xWT}2 其中yy代表类标列向量,xx代表特征行向量,WW代表回归或者分类参数矩阵 阅读全文
机器学习简史brief history of machine learning
2016-02-27 20:46 by GarfieldEr007, 524 阅读, 收藏,
摘要:
BRIEF HISTORY OF MACHINE LEARNING My subjective ML timeline (click for larger) Since the initial standpoint of science, technology and AI, scientists 阅读全文
coursera课程Text Retrieval and Search Engines之Week 4 Overview
2016-01-28 10:43 by GarfieldEr007, 308 阅读, 收藏,
摘要:
Week 4 OverviewHelp Center Week 4 On this page: Instructional Activities Time Goals and Objectives Key Phrases/Concepts Guiding Questions Readings and 阅读全文
coursera课程Text Retrieval and Search Engines之Week 3 Overview
2016-01-28 10:42 by GarfieldEr007, 335 阅读, 收藏,
摘要:
Week 3 OverviewHelp Center Week 3 On this page: Instructional Activities Time Goals and Objectives Key Phrases/Concepts Guiding Questions Readings and 阅读全文
coursera课程Text Retrieval and Search Engines之Week 2 Overview
2016-01-28 10:41 by GarfieldEr007, 245 阅读, 收藏,
摘要:
Week 2 OverviewHelp Center Week 2 On this page: Instructional Activities Time Goals and Objectives Key Phrases/Concepts Guiding Questions Readings and 阅读全文
coursera课程Text Retrieval and Search Engines之Week 1 Overview
2016-01-28 10:40 by GarfieldEr007, 353 阅读, 收藏,
摘要:
Week 1 OverviewHelp Center Week 1 On this page: Instructional Activities Time Goals and Objectives Key Phrases/Concepts Guiding Questions Readings and 阅读全文
谱聚类算法(Spectral Clustering)
2015-12-31 15:15 by GarfieldEr007, 504 阅读, 收藏,
摘要:
谱聚类(Spectral Clustering, SC)是一种基于图论的聚类方法——将带权无向图划分为两个或两个以上的最优子图,使子图内部尽量相似,而子图间距离尽量距离较远,以达到常见的聚类的目的。其中的最优是指最优目标函数不同,可以是割边最小分割——如图1的Smallest cut(如后文的Min... 阅读全文
常用机器学习方法总结
2015-12-11 19:05 by GarfieldEr007, 450 阅读, 收藏,
摘要:
1.决策树算法 决策树是一种树形分类结构,一棵决策树由内部结点和叶子结点构成,内部结点代表一个属性(或者一组属性),该结点的孩子代表这个属性的不同取 值;叶子结点表示一个类标。决策树保证每一个实例都能被一条从根结点到叶子结点的路径覆盖,叶子结点就是这条实例对应的类别,遍历这条路径的过程就是对这 条... 阅读全文
浙公网安备 33010602011771号