NoSQL DB,Which One you need?

数据库之Redis运维手册(私人整理) 1、redis配置文件 #cat 8000.conf daemonize no #[yes/no]默认值no,该参数用于定制redis服务是否以守护模式运行。 pidfile /var/run/redis.pid #默认值/var/run/redis.pid,指定redis服务的进程号文件路径,以守护模式运行时需要配置本参数; port 8000 #默认值6379,指定redis服务... 阅读详情

AllegroGraph

公司/组织:Franz Inc.
类型:Graph
简介:Modern, high performance, persistent graph database.
存储方案:Disk based, meta-data and data triples.
API(s):SPARQL, Prolog

BerkleyDB

公司/组织:Oracle
类型:Key/Value
简介:C language embeddable library for enterprise-grade, concurrent,
transactional storage services. Thread safe to avoid data corruption or
loss
存储方案:B-tree, hash table, persistent queue
API(s):C, C++ and Java
备注:Use BerkleyDB XML layer on top of BerkleyDB for XML based
applications. Comparison
of BerkleyDB and relational databases

BigTable

公司/组织:Google
类型:Sparse, distributed, persistent multidimensional sorted map.
简介:Distributed storage system for structured data. Data model provides
dynamic control over data layout and format. Data can live in memory or
on disk.
存储方案:Data is stored as an uninterpreted array of bytes. Client
applications can create structured and semi-structured data inside the
byte arrays.
API(s):Python, GQL, Sawzall API, REST, various.
备注:Overview: Bigtable: A
Distributed Storage System for Structured Data
 (PDF format)

Cassandra

公司/组织:Apache
类型:Dimensional hash table
简介:Highly scalable distributed database. CombinesDynamo’s distributed design and Bigtable’s column
family data model.
存储方案:Clusters of multiple keyspaces. The keyspace is a name space for
column families. Columns are comprised of a name, value and timestamp.
API(s):Java, Ruby, perl, Python, C#, Thrift framework.
备注:Open sourced by Facebook in 2008. Wiki, FAQ,Examples

CouchDB

公司/组织:Apache
类型:Document
简介:Distributed database with incremental replication, bi-directional
conflict detection and management.
存储方案:Ad-hoc and schema-free with a flat address space.
API(s):RESTful JSON API. JavaScript query language.
备注:CouchDB Introduction, Technical
Overview

db4o

公司/组织:Versant
类型:Object
简介:Java and .NET dual license (commercial and open source) object
database.
存储方案:Data objects are stored in the way they are defined in the
application.
API(s):Java, .NET languages.
备注:db4o db4o
database runtime engine
, about db4o

Dovetaildb

公司/组织:Millstone Creative Works
类型:JSON-based
简介:Schemaless database similar to Amazon’s SimpleDB. Open source,
standalone Java application server.
存储方案:JSON data format, “bags” (similar to tables).
API(s):HTTP and Javascript APIs
备注:Dovetaildb JavaScript
API 
reference manual

Dynomite

公司/组织:Cliff Moon
类型:Key/Value
简介:Open source Amazon
Dynamo
 clone written in Erlang.
存储方案:Distributed key/valve store, Pluggable storage engines.
API(s):Thrift API
备注:Dynomite Wiki

eXtreme
Scale

公司/组织:IBM
类型:In-memory grid/cache
简介:Distributed cache processes, partitions, replicates and manages data
across servers.
存储方案:Data and database cache, “near cache” for local subset of data. Java
persistent cache. Map reduce support.
API(s):Java APIs, REST data service
备注:eXtreme Scale Document
library
 web site

GT.M

公司/组织:FIS
类型:Hierarchical, multi-dimensional sparse arrays, content associative
memory
简介:Small footprint, multi-dimensional array with fill support for ACID
transactions, optimistic concurrency and software transactional memory.
存储方案:Unstructured array of bytes. Can be Key/Value, document oriented,
schema-less, dictionary or any other data model.
API(s):Mumps, C/C++, SQL
备注:GT.M FAQ

hamsterDB

公司/组织:Christoph Rupp
类型:Embedded storage library
简介:Lightweight embedded database engine. Supports on disk and in memory
databases.
存储方案:B+tree with variable length keys.
API(s):C++, Python, .NET and Java
备注:hamsterdb FAQ, examples, tutorial

HBase

公司/组织:Apache
类型:Sparse, distributed, persistent multidimensional sorted map.
简介:Open source, distributed, column-oriented, “Bigtable like” store
存储方案:Data row has a sortable row key and an arbitrary number of columns,
each containing arrays of bytes.
API(s):Java API, Thrift API, RESTful API
备注:Part of Apache Hadoop project. HBase Wiki, FAQ

Hypertable

公司/组织:Zvents Inc.
类型:Sparse, distributed, persistent multidimensional sorted map.
简介:High performance distributed data storage system designed to run on
distributed filesystems (but can run on local filesystems). Modeled
after Google Bigtable.
存储方案:Row key (primary key), column family, column qualifier, time stamp.
API(s):C++, Thrift API, HQL
备注:Hypertable Architectural
overview
, FAQ

Infinispan

公司/组织:JBoss Community
类型:Grid/Cache
简介:Scalable, highly available, peer to peer, data grid platform.
存储方案:Key/Value pair with optional expiration lifespan.
API(s):Java, PHP, Python, Ruby, C
备注:Infinispan FAQ, Wiki

InfoGrid

公司/组织:
类型:Graph
简介:Internet graph database made up on nodes and edges. Supports
in-memory and persistent storage alternatives including RDBMS, file
system, file grid, and custom storage.
存储方案:Nodes (meshobjects) and edges (relationships). Meshobjects can have
entity types, properties and participage in relationships. MeshObjects
raise events.
API(s):RESTful web services.
备注:InfoGrid Overview, FAQ

Keyspace

公司/组织:Scalien
类型:Key/Value
简介:Distributed (master/slave) key-value data store delivering strong
consistency, fault-tolerance and high availability.
存储方案:Uses BErkeleyDB library for For local storage. Key/Value pairs and
their state are replicated to multiple servers.
API(s):C/C++, Python, PHP, HTTP
备注:Keyspace Overview, FAQ

MemcachedDB

公司/组织:
类型:Key/Value
简介:High performance, high realiability persistent storage engine for
key/value object storage.
存储方案:Uses BerkeleyDB as storage library/backend.
API(s):Memcache protocol, C, Python, Java, perl
备注:MemcacheDB complete
guide
 (PDF format)

Mnesia

公司/组织:Ericsson
类型:Key/Value
简介:Multiuser distributed database including support for replication and
dynamic reconfiguration.
存储方案:Organized as a set of tables made up of Erlang records. Tables also
have properties including type location, persistence, etc.
API(s):Erlang
备注:Mnesia Reference
manual

MongoDB

公司/组织:10gen
类型:Document
简介:Scalable, high-performance, open source, schema-free,
document-oriented database
存储方案:JSON-like data schemas, Dynamic queries, Indexing, replication,
MapReduc
API(s):C,C++, Java, JavaScript, perl, PHP, Python, Ruby, C#, Erlang, Go,
Groovy, Haskell, Scala, F#
备注:MongoDB Documentation
Index

Neo4J

公司/组织:Neo Technology
类型:Graph
简介:Embedded, small footprint, disk based, transactional graph database
written in Java. Dual license – free and commercial.
存储方案:Graph-oriented data model with nodes, relationships and properties.
API(s):Java, Python, Ruby, Scala, Groovy, PHP, RESTful API.
备注:Neo4J Wiki, API, FAQ

Redis

公司/组织:
类型:Key/Value
简介:Key/Value store with the dataset kept in memory and saved to disk
asynchronously. “not just another key-value DB”
存储方案:Values can be strings, lists sets and sorted sets.
API(s):Python, Ruby, PHP, Erlang, Lua, C, C#, Java, Scala, perl
备注:Redis Wiki

SimpleDB

公司/组织:Amazon
类型:Item/Attribute/Value
简介:Scalable Web Service providing data storage, query and indexing in
Amazon’s cloud.
存储方案:Items (like rows of data), Attributes (like column headers), and
Values (can be multiple values)
API(s):SOAP, REST
备注:SimpleDB FAQ, Getting
Started Guide
, Developer
Guide
, API

Tokyo Cabinet

公司/组织:Mikio Hirabayashi
类型:Key/Value
简介:Library (written in C) of functions for managing files of key/value
pairs. Multi-thread support.
存储方案:Keys and Values can have variable byte length. Binary data and
strings can be used as a key and a value.
API(s):C, perl, Ruby, Java, Lua.
备注:Tokyo Cabinet Specifications, presentation(PDF
format). Also available: Tokyo Tyrant (remote service), Tokyo Distopia
(full text search), Tokyo Promenade (content management).

Voldemort

公司/组织:LinkedIn
类型:Hash Table
简介:“It is basically just a big, distributed, persistent, fault-tolerant
hash table.” High performance and availability.
存储方案:Each key is unique to a store. Each key can have at most one value.
Supported types: JSON, string, identity, protobuf, java-serialization.
API(s):Java, C++, custom clients
备注:Project Voldemort Wiki, Client
how-to

AI应用开发实战:从传统App到智能体编排的技术演进 “AI将会取代90%的app。”——这句话最近在开发者圈子里流传甚广,听起来像是一个耸人听闻的标题党,但如果你仔细拆解一下,会发现它背后指向的,是一个正在发生的、深刻的技术范式转移。作为一名开发者,我们真正关心的不是“取代”这个充满争议的词,而是: AI到底在如何重构应用形态?我们过去熟悉的“一个功能一个App”的开发模式,是否正在被颠覆?以及,作为开发者,我们的技能栈和职业路径需要做哪些调整? 很多人看到“AI取代App”的第一反应是:我的App要失业了。但更准确的判断是: AI正在将“功能”从“应用”中 阅读详情

相关推荐

Pandas入门教程(二十二):股票数据分析实战——yfinance 获取数据、技术指标与收益率分析

把 Pandas 用在真实场景,才能体会它的价值。本文以茅台股票为例,用 yfinance 获取历史行情,完成缺失值处理与列筛选,绘制收盘价曲线,计算 SMA 移动平均线与 RSI 相对强弱指数,并通过日收益率、累计收益率和波动率完成完整的量化分析流程。

Smell_of_earth的博客 305

Day16-非关系型数据库服务-redis

Redis是数百万开发人员使用的开源内存数据存储服务,经常被用于数据库、缓存、数据流引擎、以及消息中间件;简单理解:Redis 是完全开源的,遵守BSD协议,是一个高性能的key-value数据库;对于关系型数据库主要负责核心业务数据的存储和查询;对于非关系性数据库主要作为配合数据库存储和查询的缓存业务;对于非关系性数据库也是存在很多种类的:Redis(缓存型)、MogoDB(文档型)、ES(日志型)…缓存服务产品-Redis。

苦难带不走梦想 2203

【信息科学与工程学】计算机科学与自动化——第一百五十五篇 缓存系统设计03

编号366-380共15个条目,涵盖了SQL/NoSQL数据库、小文件、对象、向量、CDN、应用框架等各类数据对CPU、GPU、ASIC、FPGA、NPU、内存、SSD、HDD、RAID卡、操作系统、Redis、容器、MPP数据库、CDN等缓存硬件和软件的配置需求。:mark_cache_size,uncompressed_cache_size,merge_tree配置。:调整vm.dirty_ratio,增大vfs_cache_pressure,启用透明大页。

weixin_49199313的博客 226

MongoDB数据库设计中6条重要的经验法则

By William Zola, Lead Technical Support Engineer at MongoDB “我有丰富的sql使用经验,但是我是个MongoDB的初学者。我应该如何在MongoDB中针对一对多关系进行建模?”这是我被问及最多的问题之一。 我没法简单的给出答案,因为这有很多方案去实现。接下来我会教导你如何针对一对多进行建模。 这个话题有很多内容需要讨论,

B439984445的专栏 953

【信息科学与工程学】计算机科学与自动化——第三百零五篇 计算机系统中的各类运算操作01

编号类型领域运算操作运算类型算法算法的详细设计及数学分析参数列表及参数的数值设计应用场景 (100+条)集合运算​1集合运算关系代数并集 (Union)集合运算归并排序去重合并两个集合,去除重复元素。时间复杂度 O(n+m),n,m为两集合大小。输入:集合A, B。输出:A∪B。数据集成、多表结果合并、去重后的全量数据提取。2集合运算关系代数交集 (Intersection)集合运算哈希交集找出同时属于两个集合的元素。平均时间复杂度 O(min(n,m)),最坏 O(n*m)。输入:集合A, B。输出:A∩

weixin_49199313的博客 1388

基于知识图谱的旅游推荐系统

这一系统整合了当今先进的数据技术:采集携程等大型平台的景点数据,通过爬虫技术源源不断地获取最新的旅游信息。此外,我们采用Neo4j图数据库来存储和处理这些数据,结合自然语言处理技术,为你提供精准的旅游建议。:通过直观的知识图谱展示,了解各个景点之间微妙的联系,无论是文化背景还是地理方位,你都能从中发现新的旅行灵感。准备毕业设计的同学,关注我们,这是一个集成了多种前沿技术的项目,为你的学术成果锦上添花。:不再浪费时间,系统根据你的偏好和历史评分,利用协同过滤技术推荐适合的旅游目的地。

qalangtao的博客-qalangtao.com 773

Apache SeaTunnel详解与部署(版本2.3.11)

Apache SeaTunnel是一款高性能的分布式数据集成平台,专注于实时海量数据同步。它具有丰富可扩展的连接器(支持100+数据源)、批流一体化、多引擎支持(Zeta/Flink/Spark)等特点,解决了数据源多样、同步场景复杂等技术难题。相比于DataX、Sqoop等工具,SeaTunnel在部署难度、资源占用、功能支持等方面都具有优势。部署方式支持Docker和发布包两种形式,发布包部署需配置Maven镜像、安装插件等步骤,同时提供Web管理界面便于作业监控。该平台已被近百家企业应用于生产环境,社

super_mochi1 6322

CAT24C16 EEPROM的I²C页写与页读底层实现

I²C是一种广泛应用于嵌入式系统的双线同步串行通信协议,其核心在于严格的时序控制、地址寻址机制与主从应答逻辑。理解I²C总线物理层(上拉电阻、SCL/SDA电气特性)、协议层(START/STOP、ACK/NACK、7位地址格式)及外设寄存器操作(如SR1状态标志、CCR时钟分频),是实现可靠通信的基础。在实际工程中,EEPROM类器件(如CAT24C16)因其非易失性、页缓冲区结构和写保护(WP)功能,成为数据持久化关键载体;其16字节页写约束、内部写周期(5–10ms)与地址指针自动递增等特性,直接决定

weixin_42591413的博客 388

Spring5学习(1)-概要

Spring: the source for modern java 1. Spring Boot-BUILD(构建) ANYTHING Spring Boot is designed to get you up and running as quickly as possible, with minimal upfront configuration of Spring. Spring

Jalen备忘录 1077

从东南大学906被刷到调剂过程(内含个人感想)

介绍了自己在看到初试成绩,到后期复习以及调剂的全过程,当中也包含了自己的复试中的感想,希望大家理性看待。

Philo的博客 2854

转录组数据提取技巧汇总

数据整理是数据分析的前提,而数据的正确提取则是数据整理的前提。

zfyyzhys的博客 1784

并行分布式框架:Celery、rq、funboost

Celery是一个异步任务的调度工具,也可以叫做 "分布式任务队列(Distributed Task Queue)"。

freeking101的博客 5万+

PCA9548实战:从硬件连接到Linux驱动的I2C扩展全解析

本文详细解析了PCA9548 I2C多路复用器从硬件连接到Linux驱动配置的全过程。涵盖硬件引脚连接、电源去耦、地址选择等关键设计要点,以及Linux驱动加载、设备树配置和应用层编程技巧,帮助开发者高效实现I2C总线扩展。

weixin_28726391的博客 406

Linux入门攻坚——93、NoSQL数据库概览及MongoDB数据库入门

每个外模式都对应一个全局模式的映射规则,当全局模式发生变更(比如新增业务表、调整字段类型、修改表关联关系)时,只需要修改这个映像的映射规则,上层用户使用的外模式完全不需要改动,基于外模式开发的应用程序也无需任何修改,就可以正常运行。关系模型中,实体以及关系都是以二维表形式表示。NoSQL的兴起,是因大数据与分布式系统而产生的,其是借助于分布式技术来应对大数据时代的数据存储与处理的技术,因为在大数据时代,大数据的4V特征,导致关系型数据库在应对大数据时能力上有点捉襟见肘,因而催生了NoSQL的兴起。

kaoa000的专栏——非淡泊无以明志,非宁静无以致远 276

MediaPipe实战:从零构建人脸关键点检测应用

本文详细介绍了如何使用Google的MediaPipe框架从零构建一个实时人脸关键点检测应用。文章深入剖析了MediaPipe的三段式检测流水线,提供了Android平台的完整实战教程,包括环境配置、模型加载、摄像头集成、关键点绘制以及性能调优指南,帮助开发者快速实现高效、跨平台的人脸AR特效与交互功能。

weixin_29100927的博客 341

RK3588设备树调试实战:解决NPU不加载、USB3.0识别异常与DP显示问题

本文深入探讨了RK3588设备树调试中的三大典型问题:NPU不加载、USB3.0识别异常与DP显示问题。通过详细的问题定位步骤、关键设备树修改和调试技巧,帮助开发者快速解决硬件适配中的挑战,提升RK3588平台的开发效率。

weixin_30576827的博客 476

AI全栈实战 | 3.9-01 爬虫入门:requests 加 BeautifulSoup 三分钟跑通,但 90% 的人栽在这 4 个“礼貌“问题上

⚠️:本文内容仅用于。请遵守目标站点的服务条款与相关法律法规,不得对目标服务造成负担,不得采集、传播个人隐私数据。技术是中性的,边界在你手里。爬虫的入门技术(requests 发请求 + BeautifulSoup 解析)三分钟就能学会。真正区分"会用爬虫"和"会做爬虫的人"的,是这篇我们跑通"请求→解析→存储"的最小闭环,然后重点把"合规与礼貌"这根弦给你绷上。

源码获取 | 运行部署 | 定制二开修改,如果回复不及时请耐心等待,请见谅。 395

C#上位机之海康相机开发(SDK)实战教程

SDK的接口设计遵循“句柄-操作-释放”模式,掌握→→→的流程,就能完成基础开发;工业场景需重点关注稳定性:断连重连、内存管理、抗干扰(网口相机的IP设置、触发信号的可靠性);进阶功能(如图像处理、多相机同步)需结合视觉库(EmguCV)和硬件特性(GPIO触发),但前提是保证基础采集的稳定。建议新手先跑通SDK自带的C#示例,理解每个函数的作用,再逐步替换为自己的代码。

这里只有干货:从 Modbus 通信到 MES 对接,从 YOLO 训练到工控机部署,带你搞定工业软件开发全流程。 2192

从一段视频到可漫游的 4D 有声场景:思路拆解与最小复现手记

只需一段手机视频,你就能生成可"走进去"的 4D 有声场景:转头时狗叫声方向随之变化,走近时声音更清晰。本文拆解 Video-to-4DAVScene 的核心思路,手把手带你用几十行 Python 搭出最小管线,把视频中的运动物体变成带空间位置的声源,并附上三个真实踩坑的排障经验,代码可直接跑。🎧

helloagent的博客 232

IMSL CNL 7.0.0 (x86-64)Crack

IMSL CNL 7.0.0 (x86-64)Crack intel的算法库破解文件,这是C++版本的破解文件

深度学习工具包 DeepLearningToolbox

深度学习工具包Deprecation notice.-----This toolbox is outdated and no longer maintained.There are much better tools available for deep learning than this toolbox, e.g. [Theano](http://deeplearning.net/software/theano/), [torch](http://torch.ch/) or [tensorflow](http://www.tensorflow.org/)I would suggest you use one of the tools mentioned above rather than use this toolbox.Best, Rasmus.DeepLearnToolbox================A Matlab toolbox for Deep Learning.Deep Learning is a new subfield of machine learning that focuses on learning deep hierarchical models of data.It is inspired by the human brain's apparent deep (layered, hierarchical) architecture.A good overview of the theory of Deep Learning theory is[Learning Deep Architectures for AI](http://www.iro.umontreal.ca/~bengioy/papers/ftml_book.pdf)For a more informal introduction, see the following videos by Geoffrey Hinton and Andrew Ng.* [The Next Generation of Neural Networks](http://www.youtube.com/watch?v=AyzOUbkUf3M) (Hinton, 2007)* [Recent Developments in Deep Learning](http://www.youtube.com/watch?v=VdIURAu1-aU) (Hinton, 2010)* [Unsupervised Feature Learning and Deep Learning](http://www.youtube.com/watch?v=ZmNOAtZIgIk) (Ng, 2011)If you use this toolbox in your research please cite [Prediction as a candidate for learning deep hierarchical models of data](http://www2.imm.dtu.dk/pubdb/views/publication_details.php?id=6284)```@MASTERSTHESIS\{IMM2012-06284, author = "R. B. Palm", title = "Prediction as a candidate for learning deep hierarchical models of data", year = "2012",}```Contact: rasmusbergpalm at gmail dot comDirectories included in the toolbox-----------------------------------`NN/` - A library for Feedforward Backpropagation Neural Networks`CNN/` - A library for Convolutional Neural Networks`DBN/` - A library for Deep Belief Networks`SAE/` - A library for Stacked Auto-Encoders`CAE/` - A library for Convolutional Auto-Encoders`util/` - Utility functions used by the libraries`data/` - Data used by the examples`tests/` - unit tests to verify toolbox is workingFor references on each library check REFS.mdSetup-----1. Download.2. addpath(genpath('DeepLearnToolbox'));Example: Deep Belief Network---------------------```matlabfunction test_example_DBNload mnist_uint8;train_x = double(train_x) / 255;test_x = double(test_x) / 255;train_y = double(train_y);test_y = double(test_y);%% ex1 train a 100 hidden unit RBM and visualize its weightsrand('state',0)dbn.sizes = [100];opts.numepochs = 1;opts.batchsize = 100;opts.momentum = 0;opts.alpha = 1;dbn = dbnsetup(dbn, train_x, opts);dbn = dbntrain(dbn, train_x, opts);figure; visualize(dbn.rbm{1}.W'); % Visualize the RBM weights%% ex2 train a 100-100 hidden unit DBN and use its weights to initialize a NNrand('state',0)%train dbndbn.sizes = [100 100];opts.numepochs = 1;opts.batchsize = 100;opts.momentum = 0;opts.alpha = 1;dbn = dbnsetup(dbn, train_x, opts);dbn = dbntrain(dbn, train_x, opts);%unfold dbn to nnnn = dbnunfoldtonn(dbn, 10);nn.activation_function = 'sigm';%train nnopts.numepochs = 1;opts.batchsize = 100;nn = nntrain(nn, train_x, train_y, opts);[er, bad] = nntest(nn, test_x, test_y);assert(er < 0.10, 'Too big error');```Example: Stacked Auto-Encoders---------------------```matlabfunction test_example_SAEload mnist_uint8;train_x = double(train_x)/255;test_x = double(test_x)/255;train_y = double(train_y);test_y = double(test_y);%% ex1 train a 100 hidden unit SDAE and use it to initialize a FFNN% Setup and train a stacked denoising autoencoder (SDAE)rand('state',0)sae = saesetup([784 100]);sae.ae{1}.activation_function = 'sigm';sae.ae{1}.learningRate = 1;sae.ae{1}.inputZeroMaskedFraction = 0.5;opts.numepochs = 1;opts.batchsize = 100;sae = saetrain(sae, train_x, opts);visualize(sae.ae{1}.W{1}(:,2:end)')% Use the SDAE to initialize a FFNNnn = nnsetup([784 100 10]);nn.activation_function = 'sigm';nn.learningRate = 1;nn.W{1} = sae.ae{1}.W{1};% Train the FFNNopts.numepochs = 1;opts.batchsize = 100;nn = nntrain(nn, train_x, train_y, opts);[er, bad] = nntest(nn, test_x, test_y);assert(er < 0.16, 'Too big error');```Example: Convolutional Neural Nets---------------------```matlabfunction test_example_CNNload mnist_uint8;train_x = double(reshape(train_x',28,28,60000))/255;test_x = double(reshape(test_x',28,28,10000))/255;train_y = double(train_y');test_y = double(test_y');%% ex1 Train a 6c-2s-12c-2s Convolutional neural network %will run 1 epoch in about 200 second and get around 11% error. %With 100 epochs you'll get around 1.2% errorrand('state',0)cnn.layers = { struct('type', 'i') %input layer struct('type', 'c', 'outputmaps', 6, 'kernelsize', 5) %convolution layer struct('type', 's', 'scale', 2) %sub sampling layer struct('type', 'c', 'outputmaps', 12, 'kernelsize', 5) %convolution layer struct('type', 's', 'scale', 2) %subsampling layer};cnn = cnnsetup(cnn, train_x, train_y);opts.alpha = 1;opts.batchsize = 50;opts.numepochs = 1;cnn = cnntrain(cnn, train_x, train_y, opts);[er, bad] = cnntest(cnn, test_x, test_y);%plot mean squared errorfigure; plot(cnn.rL);assert(er<0.12, 'Too big error');```Example: Neural Networks---------------------```matlabfunction test_example_NNload mnist_uint8;train_x = double(train_x) / 255;test_x = double(test_x) / 255;train_y = double(train_y);test_y = double(test_y);% normalize[train_x, mu, sigma] = zscore(train_x);test_x = normalize(test_x, mu, sigma);%% ex1 vanilla neural netrand('state',0)nn = nnsetup([784 100 10]);opts.numepochs = 1; % Number of full sweeps through dataopts.batchsize = 100; % Take a mean gradient step over this many samples[nn, L] = nntrain(nn, train_x, train_y, opts);[er, bad] = nntest(nn, test_x, test_y);assert(er < 0.08, 'Too big error');%% ex2 neural net with L2 weight decayrand('state',0)nn = nnsetup([784 100 10]);nn.weightPenaltyL2 = 1e-4; % L2 weight decayopts.numepochs = 1; % Number of full sweeps through dataopts.batchsize = 100; % Take a mean gradient step over this many samplesnn = nntrain(nn, train_x, train_y, opts);[er, bad] = nntest(nn, test_x, test_y);assert(er < 0.1, 'Too big error');%% ex3 neural net with dropoutrand('state',0)nn = nnsetup([784 100 10]);nn.dropoutFraction = 0.5; % Dropout fraction opts.numepochs = 1; % Number of full sweeps through dataopts.batchsize = 100; % Take a mean gradient step over this many samplesnn = nntrain(nn, train_x, train_y, opts);[er, bad] = nntest(nn, test_x, test_y);assert(er < 0.1, 'Too big error');%% ex4 neural net with sigmoid activation functionrand('state',0)nn = nnsetup([784 100 10]);nn.activation_function = 'sigm'; % Sigmoid activation functionnn.learningRate = 1; % Sigm require a lower learning rateopts.numepochs = 1; % Number of full sweeps through dataopts.batchsize = 100; % Take a mean gradient step over this many samplesnn = nntrain(nn, train_x, train_y, opts);[er, bad] = nntest(nn, test_x, test_y);assert(er < 0.1, 'Too big error');%% ex5 plotting functionalityrand('state',0)nn = nnsetup([784 20 10]);opts.numepochs = 5; % Number of full sweeps through datann.output = 'softmax'; % use softmax outputopts.batchsize = 1000; % Take a mean gradient step over this many samplesopts.plot = 1; % enable plottingnn = nntrain(nn, train_x, train_y, opts);[er, bad] = nntest(nn, test_x, test_y);assert(er < 0.1, 'Too big error');%% ex6 neural net with sigmoid activation and plotting of validation and training error% split training data into training and validation datavx = train_x(1:10000,:);tx = train_x(10001:end,:);vy = train_y(1:10000,:);ty = train_y(10001:end,:);rand('state',0)nn = nnsetup([784 20 10]); nn.output = 'softmax'; % use softmax outputopts.numepochs = 5; % Number of full sweeps through dataopts.batchsize = 1000; % Take a mean gradient step over this many samplesopts.plot = 1; % enable plottingnn = nntrain(nn, tx, ty, opts, vx, vy); % nntrain takes validation set as last two arguments (optionally)[er, bad] = nntest(nn, test_x, test_y);assert(er < 0.1, 'Too big error');```[![Bitdeli Badge](https://d2weczhvl823v0.cloudfront.net/rasmusbergpalm/deeplearntoolbox/trend.png)](https://bitdeli.com/free "Bitdeli Badge")

Python+Agent入门笔记(二):列表推导式、模块和异常,这些写 Agent 前得会

接上一篇《Python+Agent入门笔记(一)》。基础语法跑通之后,这篇聊几块日常写脚本绕不开的东西。还是按 JS/Go 开发者的视角写,尽量记录我实际卡过的地方。

qq_41241612的博客 430

Statistics and Machine Learning Toolbox

Statistics and Machine Learning Toolbox(带书签),MATLAB统计学习工具

【每天一个 Python 小案例30】Python 基础知识综合小项目:我的工具盒

作为系列收官,本篇把前 29 篇的知识点合成一个能直接用的命令行工具"我的工具盒":计算器(用 ast 解析表达式,不碰 eval)、单位换算与温度转换(一张换算表 + 特例分支)、记账与分类统计(Decimal + JSON 持久化)、以及用"字典当命令表"把菜单和业务逻辑解耦的主循环。文中对比了 eval 的安全隐患与 ast 白名单的做法,解释了 lambda 闭包在字典里的写法与晚绑定陷阱,并在最后给出整个系列的知识地图,把变量、数据结构、控制流、函数、类、异常、文件、标准库八块内容对应到具体篇目。

xinjiyaya的博客 419

蓝牙模块HM-10手册

蓝牙模块HM-10手册,该手册对蓝牙模块HM-10进行了详细的介绍和使用方法以及一些AT指令。

详解Python条件语句

在 Python 中,if...else 是一种条件语句,可以根据给定的条件执行不同的操作。这个语句通常用于控制程序的流程。

2601_96798737的博客 205

Python3 协程:异步编程的核心机制与实战

协程是解决I/O密集型任务等待瓶颈的高效机制,相比线程更轻量、无锁开销。从yield到async/await的演进,使异步编程语义清晰、易于理解。通过async/await可实现并发执行多个任务(如gather、create_task),显著提升网络请求、爬虫等场景性能。关键在于使用异步库(如aiohttp)替代同步库,避免阻塞事件循环。合理控制超时与并发数,才能发挥协程最大优势。

Lxn2zh的博客 159

python如何控制浏览器

Python 可以通过 Selenium、Playwright 等工具实现浏览器自动化,从打开网页到执行复杂交互操作均可脚本化完成。Selenium生态成熟

2601_96797912的博客 76

torch.compile 差异仅 1e-6,边界样本预测为何翻转

torch.compile 加速后输出与 eager 不一致,部分预测翻转。逐层对比与 graph break 分析定位到动态分支翻转根因。容差回归仅适用于无分支翻转场景,零容忍场景应禁用编译。

昨日种种,皆成今我。今日种种,皆成新我。 158
上一篇: NoSQL DB,Which One you need?
下一篇: google 开发的Mac OS X 下的软件升级框架 update-engine 介绍
老白
博客等级 码龄23年 59粉丝 106原创
评论
成就一亿技术人!
拼手气红包6.0元
还能输入1000个字符
 
 条评论被折叠 查看
添加红包

请填写红包祝福语或标题

个

红包个数最小为10个

元

红包金额最低5元

当前余额3.43元 前往充值 >
需支付:10.00元
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付元
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值