技术笔记本

转载 万字长文带你看尽深度学习中的各种卷积网络

深度学习中的各种卷积网络大家知多少？对于那些听说过却又对它们没有特别清晰的认识的小伙伴们，Kunlun Bai这篇文章非常值得一读。Kunlun Bai 是一位人工智能、机器学习、物体学以及工程学领域的研究型科学家，在本文中，他详细地介绍了 2D、3D、1x1 、转置 、空洞（扩张）、空间可分离、深度可分离、扁平化、 分组等十多种卷积网络类型。AI 科技评论编译如下。如果你曾听过深度学习的各...

转载 训练集（train)、验证集（validation）和测试集（test）

在有监督的机器学习中，经常会说到训练集（train)、验证集（validation）和测试集（test），这三个集合的区分可能会让人糊涂，特别是，有些读者搞不清楚验证集和测试集有什么区别。划分 #如果我们自己已经有了一个大的标注数据集，想要完成一个有监督模型的测试，那么通常使用均匀随机抽样的方式，将数据集划分为训练集、验证集、测试集，这三个集合不能有交集，常见的比例是8:1:1，当然比例是...

转载 SIFT: Theory and Practice - Generating a feature（转）

SIFT: Theory and Practice Generating a featureSeries: SIFT: Theory and Practice:Introduction The scale space LoG approximations Finding keypoints Getting rid of low contrast keypoints Key...

转载 SIFT: Theory and Practice - Keypoint orientations（转）

SIFT: Theory and Practice Keypoint orientationsSeries: SIFT: Theory and Practice:Introduction The scale space LoG approximations Finding keypoints Getting rid of low contrast keypoints Ke...

转载 SIFT: Theory and Practice - Getting rid of low contrast keypoints（转）

SIFT: Theory and Practice Getting rid of low contrast keypointsSeries: SIFT: Theory and Practice:Introduction The scale space LoG approximations Finding keypoints Getting rid of low contra...

转载 SIFT: Theory and Practice - Finding keypoints（转）

SIFT: Theory and Practice Finding keypointsSeries: SIFT: Theory and Practice:Introduction The scale space LoG approximations Finding keypoints Getting rid of low contrast keypoints Keypoi...

转载 SIFT: Theory and Practice - LoG approximations （转）

SIFT: Theory and Practice LoG approximationsSeries: SIFT: Theory and Practice:Introduction The scale space LoG approximations Finding keypoints Getting rid of low contrast keypoints Keypo...

转载 SIFT: Theory and Practice - The scale space（转）

SIFT: Theory and Practice The scale spaceSeries: SIFT: Theory and Practice:Introduction The scale space LoG approximations Finding keypoints Getting rid of low contrast keypoints Keypoint...

转载 SIFT: Theory and Practice - Introduction (转)

SIFT: Theory and PracticeLearn how the famous SIFT keypoint detector works in the background. This paper led a mini revolution in the world of computer vision!IntroductionSeries: SIFT: Theory an...

转载 A tutorial on binary descriptors – part 5 – The FREAK descriptor（转）

A tutorial on binary descriptors – part 5 – The FREAK descriptorThis is our fifth post in the series about binary descriptors and here we will talk about the FREAK[4] descriptor. This is the last de...

转载 A tutorial on binary descriptors – part 4 – The BRISK descriptor（转）

A tutorial on binary descriptors – part 4 – The BRISK descriptorThis fourth post in our series about binary descriptors that will talk about the BRISK descriptor [1]. We had an introduction to patch...

转载 A tutorial on binary descriptors – part 3 – The ORB descriptor（转）

A tutorial on binary descriptors – part 3 – The ORB descriptorThis third post in our series about binary descriptors that will talk about the ORB descriptor [1]. We had an introduction to patch desc...

转载 A tutorial on binary descriptors – part 2 – The BRIEF descriptor（转）

A tutorial on binary descriptors – part 2 – The BRIEF descriptorFollowing the previous posts that provided both an introduction to patch descriptors in general and specifically to binary descriptors...

转载 Tutorial on Binary Descriptors – part 1(转)

Tutorial on Binary Descriptors – part 1Why Binary Descriptors?Following the previous post on descriptors, we’re now familiar with histogram of gradients (HOG) based patch descriptors. SIFT[1], SUR...

转载 How to do distributed locking (转)

As part of the research for my book, I came across an algorithm called Redlock on the Redis website. The algorithm claims to implement fault-tolerant distributed locks (or rather, leases [1]) on top o...

原创 生成RSA私钥和公钥

#私钥openssl genrsa -out rsakey0.pem 1024#公钥openssl rsa -in rsakey0.pem -pubout -out rsakey0-pub.pem#pkcs8格式私钥openssl pkcs8 -topk8 -in rsakey0.pem -out pkcs8_rsa_private_key.pem -nocrypt

转载 Spring Beginners to Professionals

http://techieme.in/category/j2ee-concepts/spring-beginners-to-professionals/?order=ascSpring Beginners to Professionals

转载 Understanding reference in java

http://www.pawlan.com/monica/articles/refobjs/https://stackoverflow.com/questions/299659/whats-the-difference-between-softreference-and-weakreference-in-javaAnalogy: Assume a JVM is a kingdo

原创 配置MongoDB Sharding

1) 启动副本集三个shard，每个shard三个副本，三台机器（170/171/172），每台机器分别对应三个shard中的一个副本#170mongod --shardsvr --replSet shard-a --dbpath /mongo-data/rs-170-a --port 30000 --logpath /mongo-data/rs-170-a.log --fork --no

原创 Recurrent Neural Networks - collections

The Unreasonable Effectiveness of Recurrent Neural NetworksUnderstanding LSTM NetworksRECURRENT NEURAL NETWORKREGULARIZATIONNeural Networks and Deep Learning

转载 AOP技术基础

1、引言2、AOP技术基础3、Java平台AOP技术研究4、.Net平台AOP技术研究2.1 AOP技术起源AOP技术的诞生并不算晚，早在1990年开始，来自Xerox Palo Alto Research Lab（即PARC）的研究人员就对面向对象思想的局限性进行了分析。他们研究出了一种新的编程思想，借助这一思想或许可以通过减少代码重复模块从而帮助开发人员

原创 Lp距离, L1范数, 和L2范数

LpL_p距离, L1L_1范数, 和 L2L_2范数LpL_p 距离定义为: Lp(xi,xj)=(∑l=1n|x(l)i−x(l)j|p)1p L_p(x_i, x_j) = \left( \sum_{l=1}^{n} | x_i^{(l)} - x_j^{(l)} |^p \right)^{\frac{1}{p}} 其中xi∈Rnx_i \in \mathbf{R}^n,

原创 TensorFlow - ImportError: cannot import name pywrap_tensorflow

>>> import tensorflow as tfTraceback (most recent call last):  File "", line 1, in   File "tensorflow/__init__.py", line 23, in     from tensorflow.python import *  File "tensorflow/python/_

原创 Yann LeCun, Geoffrey E. Hinton, and Yoshua Bengio

原创 深度学习历史图

原创 机器学习资源列表

Machine Learning Group at Department of Engineering, University of CambridgeMachine Learning Group at University of columbiaMachine Learning Research Group at university of texas at austinMachin

原创 大数据资源收集

数据可视化https://github.com/showcases/data-visualization

转载 Understanding the Internal Message Buffers of Storm

Understanding the Internal Message Buffers of StormJun 21st, 2013 Table of ContentsInternal messaging within Storm worker processesIllustrationDetailed descriptionWorker processesExecutors

原创 shell使用随笔

001 对文件某一列求和awk '{sum += $collum};END {print sum}' /path/to/your/file

转载 Shannon entropy

Shannon entropy is one of the most important metrics in information theory. Entropy measures the uncertainty associated with a random variable, i.e. the expected value of the information in the messag

转载 Positive-definite kernel

http://www.encyclopediaofmath.org/index.php/Bounded_operator

转载 Hilbert space

DefinitionA Hilbert space H is a real or complex inner product space that is also a complete metric space with respect to the distance function induced by the inner product.[2] To say that

转载 Positive-definite matrix

In linear algebra, a symmetric n × n real matrix M is said to be positive definite if zTMz is positive for every non-zero columnvector z of n real numbers. Here zT denotes thetranspose

原创 MVN使用随笔

001 创建项目mvn archetype:generate -DgroupId=com.funshion.push.monitor -DartifactId=push-monitor -DarchetypeArtifactId=maven-archetype-quickstart -DinteractiveMode=false002 [WARNING] File encoding h

转载 The storage wars: Shadow Paging, Log Structured Merge and Write Ahead Logging

The storage wars: Shadow Paging, Log Structured Merge and Write Ahead Loggingprevious: Seek, and you shall find, or maybe delay a lot next: New interview questionI’ve been doing a lo

转载 MySQL删除表数据

在MySQL中有两种方法可以删除数据，一种是DELETE语句，另一种是TRUNCATE TABLE语句。DELETE语句可以通过WHERE对要删除的记录进行选择。而使用TRUNCATE TABLE将删除表中的所有记录。因此，DELETE语句更灵活。　　如果要清空表中的所有记录，可以使用下面的两种方法：　　DELETE FROM table1　　TRUNCATE TABLE table1

转载 MySQL存储过程之事务管理

原文链接：http://hideto.iteye.com/blog/195275MySQL存储过程之事务管理ACID:Atomic、Consistent、Isolated、Durable存储程序提供了一个绝佳的机制来定义、封装和管理事务。1，MySQL的事务支持MySQL的事务支持不是绑定在MySQL服务器本身，而是与存储引擎相关：Java代码  

转载 MySQL事务隔离级别详解

SQL标准定义了4类隔离级别，包括了一些具体规则，用来限定事务内外的哪些改变是可见的，哪些是不可见的。低级别的隔离级一般支持更高的并发处理，并拥有更低的系统开销。Read Uncommitted（读取未提交内容）       在该隔离级别，所有事务都可以看到其他未提交事务的执行结果。本隔离级别很少用于实际应用，因为它的性能也不比其他级别好多少。读取未提交的数据，也被称之为脏读（Dirty

原创 MySQL使用随笔

#查看版本：mysql --versionselect version();status;

转载 HBase High Level Architecutre