新起点

原创 Amortized Analysis Examples

Amortized analysis refers to determining the time-averaged running time for a sequence of operations.  It is different from what is commonly referred to as average case analysis, because amortized ana...

原创 Properties of B-Tree

1) All leaves are at same level.2) A B-Tree is defined by the term minimum degree ‘t’. The value of t depends upon disk block size.3) Every node except root must contain at least t-1 keys. Root may co...

转载 主成分分析（二）

转自：https://www.cnblogs.com/jerrylead/archive/2011/04/18/2020216.html主成分分析（Principal components analysis）-最小平方误差解释 接上篇3.2 最小平方误差理论          假设有这样的二维样本点（红色点），回顾我们前面探讨的是求一条直线，使得样本点投影到直线上的点的方差最大...

转载 主成分分析(一)

转自：https://www.cnblogs.com/jerrylead/archive/2011/04/18/2020209.html主成分分析（Principal components analysis）-最大方差解释     在这一篇之前的内容是《Factor Analysis》，由于非常理论，打算学完整个课程后再写。在写这篇之前，我阅读了PCA、SVD和LDA。这几个模型相近，却都...

转载 线性判别分析（二）

转自：https://www.cnblogs.com/jerrylead/archive/2011/04/21/2024389.html（Linear Discriminant Analysis）（二）4. 实例      将3维空间上的球体样本点投影到二维上，W1相比W2能够获得更好的分离效果。            PCA与LDA的降维对比：            ...

转载 线性判别分析(一)

转自：https://www.cnblogs.com/jerrylead/archive/2011/04/21/2024384.html线性判别分析（Linear Discriminant Analysis）（一）1. 问题     之前我们讨论的PCA、ICA也好，对样本数据来言，可以是没有类别标签y的。回想我们做回归时，如果特征太多，那么会产生不相关特征引入、过度拟合等问题。我们可...

原创 Lazy learning algorithm VS Eager learning algorithm

1.Lazy:     a.Defers data processing until it recieves a request to classify an unlabeled example.     b.Replies to a request for information by combining its stored training data.     c.Discard...

原创 KNN——A few thoughts on picking a value for "K"

1.We just need to try out a few values before settling on one.2.Low values of K can be noisy and subject to the effects of outliers.3.Large values of K smooth over thins, but if k is very large, a...

原创 PCA vs LDA

1. The differences between PCA and LDA:    a.PCA has a bad discriminability between different classes, but it has  high variance in one class.    b.LDA has a good discriminability between differen...

原创 AI midterm exam review

1.State-space    components of state-space:         state variableds   state    state space   operator set    startd state   goal state   state space search graph(tree)      There are two kind...

原创 Properties Of Agents

Autonomous - is capable acting without direct external intervention. It has some degree of control over its internal state and actions based on its own experiences.Interactive - communicates with th...

原创 Sorting Summary

原创 How to Find Complexity?

原创 Necessary Proof Techniques

1.Direct Proof – combining axioms, definitions, and earlier theorems.2.Proof by Induction – theorem holds for n; show that it holds for n+1.3.Disprove by Counter Example.4.Proof by Contradiction – ...

原创 Recurrence

1. Three different solutions:    a.Substitution Method.    b.Recursion-Tree Method.    c.Master Method.2.Substitution Method:Guess the form of the solution, then us mathematical induction to s...

原创 Dynamic Programming

1. two fields to use dynamic programming         a. find a solution with the optimal value         b.find the maximum and minimum2.Unlike divide and conquer, the subproblems of the dynamic progr...

原创 The typical steps of greedy algorithm

1.Cast the optimization problem as one in which we make a choice and are left with one subproblem to solve.2.Prove that there is always an optimal solution that makes the greedy choice, so that the ...

转载 differences between dynamic, divide-and-conquer, and greedy programming

Divide-and-Conquer:Strategy: Break a small problem into smaller sub-problems. Smaller sub-problems will most likely be a scaled down version of the original problem. Solve the smaller problems. Combin...

原创 The step of CFG convert to Chomsky normal form

If there is a S on the right, add a new start S0. Remove ∈  rules. Remove unit rules. Add additional variables and rules.

原创 Python网络数据采集读书笔记

1.网络爬虫   • 通过网站域名获取 HTML 数据   • 根据目标信息解析数据   • 存储目标信息   • 如果有必要，移动到另一个网页重复这个过程2.每当你调用 BeautifulSoup 对象里的一个标签时，增加一个检查条件保证标签确实存在是很聪明的做法。3.在写爬虫的时候，思考代码的总体格局，让代码既可以捕捉异常又容易阅读，这是很重要的。如果你还希望能够很大程度地重...

原创 Requset 学习笔记

1.request.raise_for_status()This is a new function to add to our existing toolbox of ways to check the status of a request. When you call the function r.raise_for_status(), Requests will check the s...

转载 covariance和correlation的区别

covariance是计量经济中的协变差或称协方差，而correlation是指两个数值的相关性。1、covariance（协变）：计量经济中的协变差或称协方差；2、correlation（相关性）：指两个数值的相关性，取值一般在-1和+1之间，取0表示不相关，取-1表示负相关，取+1表示正相关。3、两者的区别如下：（1）协方差是一个用于测量投资组合中某一具体投资项目相对于另一投资...

转载 Let's implement a Gaussian Naive Bayes classifier in Python

  A typical example of problem ML tries to solve is classification. It can be expressed as the ability, given some input data, to assign a ‘class label’ to a sample.To make things clearer, let’s...

转载 2 ways of using Naive Bayes classification for numeric attributes

Previously we described how Naive Bayes classification works using the classic golf example. There we mentioned that the algorithm is designed to work primarily with categorical attributes such as Out...

转载 Naive Bayes Classification With Sklearn

This tutorial details Naive Bayes classifier algorithm, its principle, pros & cons, and provides an example using the Sklearn python Library.ContextLet’s take the famous Titanic Disaster datas...

原创 Three types of Naïve Bayes classifiers

There are three types of Naïve Bayes classifiers. When handling real-time data with continuous distribution, Naïve Bayes classifier considers that the big data is generated through a Gaussian process ...

转载 Better Naive Bayes: 12 Tips To Get The Most From The Naive Bayes Algorithm

1. Missing DataNaive Bayes can handle missing data.Attributes are handled separately by the algorithm at both model construction time and prediction time.As such, if a data instance has a missin...