基础概念

深度学习和人工智能

开源深度学习框架

此处个人选择 Deeplearning4j。用什么框架区别并不大，静下心来学习原理才是重要的。

Deeplearning4j

Deeplearning4j is the first commercial-grade, open-source, distributed deep-learning library written for Java and Scala.

Integrated with Hadoop and Spark, DL4J is designed to be used in business environments on distributed GPUs and CPUs.

可应用于以下深度学习领域：

  • 人脸/图像识别

  • 语音搜索

  • 语音转文字（Speech to text）

  • 垃圾信息过滤（异常侦测）

  • 电商欺诈侦测

Quick Start

官网地址

Prerequisites

  1. Java (developer version) 1.7 or later (Only 64-Bit versions supported)

  2. Apache Maven (automated build and dependency manager)

  3. Git

  4. IntelliJ IDEA or Eclipse

本机测试

  • Jdk
houbinbindeMacBook-Pro:003_condition houbinbin$ java -version
java version "1.8.0_91"
Java(TM) SE Runtime Environment (build 1.8.0_91-b14)
Java HotSpot(TM) 64-Bit Server VM (build 25.91-b14, mixed mode)
  • Maven
houbinbindeMacBook-Pro:003_condition houbinbin$ mvn -v
Apache Maven 3.3.9 (bb52d8502b132ec0a5a3f4c09453c07478323dc5; 2015-11-11T00:41:47+08:00)
Maven home: /usr/local/maven/maven3.3.9
Java version: 1.8.0_91, vendor: Oracle Corporation
Java home: /Library/Java/JavaVirtualMachines/jdk1.8.0_91.jdk/Contents/Home/jre
Default locale: zh_CN, platform encoding: UTF-8
OS name: "mac os x", version: "10.11.3", arch: "x86_64", family: "mac"
  • Git
houbinbindeMacBook-Pro:003_condition houbinbin$ git --version
git version 2.8.1
  • IDE

本机使用IDEA。

DL4J Examples

一、下载和安装

$ git clone https://github.com/deeplearning4j/dl4j-examples.git
$ cd dl4j-examples/
$ mvn clean install
  • Download

发现命令行安装比较慢。就直接下载。

  • pom.xml

To run DL4J in your own projects, we highly recommend using Maven for Java users, or a tool such as SBT for Scala. The basic set of dependencies and their versions are shown below. This includes:

  1. deeplearning4j-core, which contains the neural network implementations

  2. nd4j-native-platform, the CPU version of the ND4J library that powers DL4J

  3. datavec-api - Datavec is our library vectorizing and loading data

这些Demo里面已经有了，无需再次引入。

  • Install
Book-Pro:dl4j-examples houbinbin$ pwd
/Users/houbinbin/IT/learn/DL/dl4j-examples-master/dl4j-examples
houbinbindeMacBook-Pro:dl4j-examples houbinbin$ mvn clean install
[INFO] Scanning for projects...
....
[INFO] BUILD SUCCESS
[INFO] ------------------------------------------------------------------------
[INFO] Total time: 01:05 min
[INFO] Finished at: 2017-04-16T13:37:47+08:00
[INFO] Final Memory: 89M/897M
[INFO] ------------------------------------------------------------------------
  • Run Demo

根据官方推荐。运行 MLPClassifierLinear 里面的 Main() 即可。

Terminal log

o.n.l.f.Nd4jBackend - Loaded [CpuBackend] backend
o.n.n.NativeOpsHolder - Number of threads used for NativeOps: 4
o.n.n.Nd4jBlas - Number of threads used for BLAS: 4
o.n.l.a.o.e.DefaultOpExecutioner - Backend used: [CPU]; OS: [Mac OS X]
o.n.l.a.o.e.DefaultOpExecutioner - Cores: [8]; Memory: [3.6GB];
o.n.l.a.o.e.DefaultOpExecutioner - Blas vendor: [OPENBLAS]
o.d.o.l.ScoreIterationListener - Score at iteration 0 is 0.6082896423339844
...
o.d.o.l.ScoreIterationListener - Score at iteration 10 is 0.5995052337646485
Evaluate model....

Examples labeled as 0 classified by model as 0: 100 times
Examples labeled as 1 classified by model as 1: 100 times


==========================Scores========================================
 Accuracy:        1
 Precision:       1
 Recall:          1
 F1 Score:        1
========================================================================
****************Example finished********************

效果如下

MLPClassifierLinear