11 2018 档案
摘要:1.深度神经网络 1.以往神经网络采用单或双隐层结构,虽然参照了生物上的神经元,但是从本质上来说还是数学,以函数嵌套形成。 2.通常使用的激活函数是连续可微(differentiable)的,sigmoid函数或者是右侧的,本质上是减少梯度的降低速度。 3.现在神经网络的层数在逐级增加,几千层的也比
阅读全文
摘要:转自:https://wenku.baidu.com/view/ccfa573a3968011ca30091d6.html https://www.cnblogs.com/arkenstone/p/5496761.html 1.定义 Kolmogorov-Smirnov是比较一个频率分布f(x)与理
阅读全文
摘要:转自:https://zhidao.baidu.com/question/2052933434631672387.html 1.解释 解释:表达值倍数变化 ,分析,消除可能的混杂因素,必要时可以用读段 的绝对表达值倍数变化(fold-change)来作为补充。 Fold change 用于描述一个初
阅读全文
摘要:转自:https://baike.baidu.com/item/%E7%9F%A9%E9%98%B5%E6%9D%A1%E4%BB%B6%E6%95%B0/10150161?fr=aladdin https://wenku.baidu.com/view/27cc784a02768e9951e738f
阅读全文
摘要:转自:https://baike.baidu.com/item/%E5%9F%BA%E5%9B%A0%E8%A1%A8%E8%BE%BE%E8%B0%B1/364115?fr=aladdin 1.定义 gene expression profile:指通过构建处于某一特定状态下的细胞或组织的非偏性c
阅读全文
摘要:转自:https://zhidao.baidu.com/question/494249074914968332.html SVM使用拉格朗日乘子法更为高效地求解了优化问题。 SVM将寻找具有最大几何间隔划分超平面的任务转化成一个凸优化问题,如下所示: //其中,“s.t.”,指 subject to
阅读全文
摘要:转自:https://wenku.baidu.com/view/3815adfdfad6195f302ba6c0.html 1.约束条件下多变量的优化方法 2.等式约束下的拉格朗日乘子法 2.1等式约束下极值存在的必要条件 上述两页讲了两个必要条件,无约束条件下和等式约束下存在极小的必要条件。 2.
阅读全文
摘要:T检验P值表格来自:https://blog.csdn.net/m0_37777649/article/details/74937242 Z检验表格来自:https://wenku.baidu.com/view/4926021f10a6f524ccbf85d6.html 确定Z检验结果转自:http
阅读全文
摘要:转自:https://blog.csdn.net/m0_37777649/article/details/74937242 1.什么是T检验? T检验是假设检验的一种,又叫student t检验(Student’s t test),主要用于样本含量较小(例如n<30),总体标准差σ未知的正态分布资料
阅读全文
摘要:转自:https://blog.csdn.net/Orange_Spotty_Cat/article/details/80312154 1.意义 Z-Score通过(x-μ)/σ将两组或多组数据转化为无单位的Z-Score分值,使得数据标准统一化,提高了数据可比性,削弱了数据解释性。 就是减去均值,
阅读全文
摘要:转自:https://www.zhihu.com/question/19734616/answer/349132554 https://www.cnblogs.com/zhangchaoyang/articles/2631907.html 1.Pearson 分母是两个变量的标准差,分子是协方差。
阅读全文
摘要:转自:http://blog.sina.com.cn/s/blog_40d4ae110101fjzy.html 1 二代测序与基因芯片的区别与优缺点。 生物芯片相对第二代测序而言,优势在于价格便宜,便于分析。缺点则在于必须有参考序列(因为生物芯片的探针设计就是根据参考序列设计的)。 相同点: 高通量
阅读全文
摘要:转自:https://www.cnblogs.com/ysugyl/p/8711205.html Grid Search:一种调参手段;穷举搜索:在所有候选的参数选择中,通过循环遍历,尝试每一种可能性,表现最好的参数就是最终的结果。其原理就像是在数组里找最大值。(为什么叫网格搜索?以有两个参数的模型
阅读全文
摘要:转自:https://blog.csdn.net/dang_boy/article/details/78504258 https://www.cnblogs.com/Belter/p/8536939.html https://www.cnblogs.com/Belter/p/8536939.html
阅读全文
摘要:转自:https://www.jianshu.com/p/85df3a5e2eb4 http://www.cnblogs.com/aloiswei/p/5928748.html 1.多重共线性 如果存在某个常数C0,C1,C2,并且使得线性等式满足以下情况: 则X1,X2是精准共线性的。 对于P(>
阅读全文
摘要:paper:《肿瘤基因表达谱分类特征基因选取问题及分析方法研究》 1.肿瘤基因表达谱 数据的一个显著特点是样本维数过高,每个样本都记录了组织细胞中所有可测基因的表达水平,但实际上只有少数基因才真正同样本类别相关,包含了样本分类信息,这些基因被称为分类特征基因. 2.肿瘤基因表达谱是指利用DNA芯片所
阅读全文
摘要:转自:https://www.jianshu.com/p/43318a3dc715?from=timeline&isappinstalled=0 https://blog.csdn.net/ericcchen/article/details/72357411 实际含义包括数据和假设的概率模型之间的差
阅读全文
摘要:Suppose that all the keys in a binary tree are distinct positive integers. Given the preorder and inorder traversal sequences, you are supposed to out
阅读全文
摘要:Given any string of N (≥5) characters, you are asked to form the characters into the shape of U. For example, helloworld can be printed as: h d e l l
阅读全文
摘要:A clique is a subset of vertices of an undirected graph such that every two distinct vertices in the clique are adjacent. A maximal clique is a clique
阅读全文
摘要:转自:https://www.cnblogs.com/wyuzl/p/7699818.html 1.在决策树算法中,就是根据信息论的方法找到最合适的特征来划分数据集。在这里,我们首先要计算所有类别的所有可能值的香农熵,根据香农熵来我们按照取最大信息增益(information gain)的方法划分我
阅读全文
摘要:转自:https://www.cnblogs.com/ohshit/p/5629581.html (1)条件概率公式 设A,B是两个事件,且P(B)>0,则在事件B发生的条件下,事件A发生的条件概率(conditional probability)为: P(A|B)=P(AB)/P(B) (2)乘法
阅读全文
摘要:1100 Mars Numbers (20 分) People on Mars count their numbers with base 13: Zero on Earth is called "tret" on Mars. The numbers 1 to 12 on Earch is call
阅读全文
摘要:The ranklist of PAT is generated from the status list, which shows the scores of the submissions. This time you are supposed to generate the ranklist
阅读全文
摘要:Given a list of N student records with name, ID and grade. You are supposed to sort the records with respect to the grade in non-increasing order, and
阅读全文
摘要:Given an integer with no more than 9 digits, you are supposed to read it in the traditional Chinese way. Output Fu first if it is negative. For exampl
阅读全文
摘要:There is a kind of balanced binary search tree named red-black tree in the data structure. It has the following 5 properties: (1) Every node is either
阅读全文
摘要:Suppose that all the keys in a binary tree are distinct positive integers. A unique binary tree can be determined by a given pair of postorder and ino
阅读全文
摘要:> c [,1] [,2] [,3] [,4] [1,] 1 4 7 10 [2,] 2 5 8 11 [3,] 3 6 9 12 > eigen(c) eigen() decomposition $`values` [1] 3.620937e+01 -2.209373e+00 1.599839e-15 ...
阅读全文
摘要:British astronomer Eddington liked to ride a bike. It is said that in order to show off his skill, he has even defined an "Eddington number", E -- tha
阅读全文
摘要:About 900 years ago, a Chinese philosopher Sima Guang wrote a history book in which he talked about people's talent and virtue. According to his theor
阅读全文
摘要:Eva loves to collect coins from all over the universe, including some other planets like Mars. One day she visited a universal shopping mall which cou
阅读全文
摘要:转自:https://www.cnblogs.com/terencezhou/p/6235974.html https://www.cnblogs.com/nsnow/p/4758202.html 这个讲到了协方差矩阵的几何解释,,我现在是看不懂的。。。 2020-9-14更新—————— http
阅读全文
摘要:Look-and-say sequence is a sequence of integers as the following: D, D1, D111, D113, D11231, D112213111, ... where D is in [0, 9] except 1. The (n+1)s
阅读全文
摘要:1089 Insert or Merge (25 分) According to Wikipedia: Insertion sort iterates, consuming one input element each repetition, and growing a sorted output
阅读全文
摘要:Among all the factors of a positive integer N, there may exist several consecutive numbers. For example, 630 can be factored as 3×5×6×7, where 5, 6, a
阅读全文
摘要:Being unique is so important to people on Mars that even their lottery is designed in a unique way. The rule of winning is simple: one bets on a numbe
阅读全文
摘要:Given two sets of integers, the similarity of the sets is defined to be Nc/Nt×100%, where Nc is the number of distinct common numbers shared
阅读全文
摘要:转自:https://www.cnblogs.com/mfrbuaa/p/5319365.html 麻省理工学院线性代数课笔记。 1.一维投影 2.三维投影 最后还涉及到了最小二乘法的几何意义。也就是投影得到的公式啊!
阅读全文
摘要:The basic task is simple: given N real numbers, you are supposed to calculate their average. But what makes it complicated is that some of the input n
阅读全文
摘要:全文转自:https://www.cnblogs.com/gatherstars/p/6084696.html#commentform 这篇真的讲的清楚明白!要多复习!加深记忆! 1.概述 AUC(Area Under roc Curve)是一种用来度量分类模型好坏的一个标准。这样的标准其实有很多,
阅读全文
摘要:On a broken keyboard, some of the keys are worn out. So when you type some sentences, the characters corresponding to those keys will not appear on sc
阅读全文
摘要:The string APPAPT contains two PAT's as substrings. The first one is formed by the 2nd, the 4th, and the 6th characters, and the second one is formed
阅读全文
摘要:Given a tree, you are supposed to tell if it is a complete binary tree. Input Specification: Each input file contains one test case. For each case, th
阅读全文
摘要:Zhejiang University has 40000 students and provides 2500 courses. Now given the student name lists of all the courses, you are supposed to output the
阅读全文
摘要:Given any positive integer N, you are supposed to find all of its prime factors, and write them in the format N = p1k1×p2k2×⋯×pm
阅读全文
摘要:1.CreateSeuratObject 转自:https://cloud.tencent.com/developer/article/1055892 # Initialize the Seurat object with the raw (non-normalized data). Keep al
阅读全文
摘要:转自:https://yoyoyohamapi.gitbooks.io/mit-ml/content/%E7%89%B9%E5%BE%81%E9%99%8D%E7%BB%B4/articles/PCA.html https://www.jianshu.com/p/162bb4ea1b7f 1.有什么
阅读全文
摘要:"Let's C" is a popular and fun programming contest hosted by the College of Computer Science and Technology, Zhejiang University. Since the idea of th
阅读全文
摘要:The Japanese language is notorious for its sentence ending particles. Personal preference of such particles can be considered as a reflection of the s
阅读全文
摘要:1.R中NA,NaN,Inf代表什么? NA:缺失数据 NaN:无意义的数,比如sqrt(-2) Inf:正无穷大 -Inf:负无穷大 判断是否存在nan值 is.na(x)#是否存在nan anyNA(x) sum(is.na(x))#存在几个nan值 2.确定一个数值型vector的第一个最值(
阅读全文
摘要:This time you are asked to tell the difference between the lowest grade of all the male students and the highest grade of all the female students. Inp
阅读全文
摘要:A linked list consists of a series of structures, which are not necessarily adjacent in memory. We assume that each structure contains an integer key
阅读全文
摘要:The task of this problem is simple: insert a sequence of distinct positive integers into a hash table first. Then try to find another sequence of inte
阅读全文
摘要:在Rstdio使用的是3.5.1的64位R版本中遇到问题:The Perl script 'WriteXLS.pl' failed to run successfully. 首先使用 Sys.which("perl") 输出空串,表明未安装perl。 perl "" 安装上perl之后,运行又出现了
阅读全文
摘要:参数检验就是已知数据的精确分布模型,根据数据来求出模型中的未知参数;而非参数检验就是无需对样本总体分布(比如满足正态分布)做出假设。 1.符号检验 转自:https://baike.baidu.com/item/%E9%9D%9E%E5%8F%82%E6%95%B0%E6%A3%80%E9%AA%8
阅读全文
摘要:转自:https://baike.baidu.com/item/RPKM/1197657 均反应基因的表达水平 1.RPKM的计算公式 分母是总共比对到这个基因的reads的数目(条 为单位),分母是:比对上的reads的总数(百万条为单位);外显子的长度也就是基因的长度(KB为单位)。 2.举个计
阅读全文
摘要:Consider a positive integer N written in standard notation with k+1 digits ai as ak⋯a1a0 with 0≤ai<10for all i and ak>0. Then N is p
阅读全文