The race of tech leaders for AI is so rude and competitive. Things are getting changed day after day by AI. Web development is no exception.

AI的技术领导者竞赛是如此粗鲁和竞争。 人工智能日新月异。 Web开发也不例外。

Having a website is mandatory for all business niches in the market. And AI will empower web design and development. According to a Statista report, the revenue generation from AI is expected to reach $126 Billion in 2025.

网站是市场上所有业务细分领域的必备条件。 人工智能将支持Web设计和开发。 根据Statista的一份报告，到2025年，来自AI的收入预计将达到1260亿美元 。

As a web software engineer, I choose to give you today a brief vision about the top AI and ML frameworks for the web to keep an eye on in 2020 and beyond. This will change the web developer role and in-demand skills in the future.

作为一名网络软件工程师，我选择今天为您简要介绍一下网络的顶级AI和ML框架，以便在2020年及以后继续关注。 将来，这将改变Web开发人员的角色和需求技能。

Ready? Let’s go ♪~ ᕕ(ᐛ)ᕗ

准备？ 出发吧♪〜ᕕ(ᐛ)ᕗ

张量流 (Tensorflow)

TensorFlow is a library for machine learning in JavaScript developed by the Google Brain team and introduced in 2015.

TensorFlow是由Google Brain团队开发并于2015年推出JavaScript机器学习库。

特点： (Characteristics:)

• Written In: C++, Python and, CUDA

  撰写： C ++，Python和CUDA

• Platforms: Linux, macOS, Microsoft Windows, and Android.

  平台： Linux，macOS，Microsoft Windows和Android。

关键点： (Key points:)

• Web developers can run existing models in the browser or under Node.js.Web开发人员可以在浏览器中或在Node.js下运行现有模型。
• Retrain pre-existing Machine Learning models using your data (Transfer Learning)使用数据重新训练已有的机器学习模型(转移学习)
• Develop ML with flexible and intuitive JavaScript APIs.使用灵活直观JavaScript API开发ML。

阿帕奇·马豪(Apache Mahout) (Apache Mahout)

Apache Mahout is an open-source machine learning framework developed by Apache Software Foundation in 2008.

Apache Mahout是Apache Software Foundation在2008年开发的开源机器学习框架。

特点： (Characteristics:)

• Written In: Java and Scala

  撰写： Java和Scala

• Platforms: Apache Hadoop, Apache Spark, Apache Flink, Scala+

  平台： Apache Hadoop，Apache Spark，Apache Flink，Scala +

关键点： (Key points:)

• Improve web application performance by using distributed linear algebra frameworks with mathematically expressive R-Like Scala DSL for distributed algorithms.

  通过将分布式线性代数框架与具有数学表达能力的R-Like Scala DSL用于分布式算法一起使用，可以提高Web应用程序的性能。

• Build scalably supervised (classification) and unsupervised (grouping) machine learning systems in the cloud with Apache Hadoop library.使用Apache Hadoop库在云中构建可扩展的监督(分类)和不受监督(分组)机器学习系统。
• Clustering implementations to discover similarities between items and grouping data thus enhance recommendation systems.聚类实现以发现项目和分组数据之间的相似性，从而增强了推荐系统。

Microsoft认知工具包 (Microsoft Cognitive Toolkit)

A free and easy-to-use open-source and deep-learning toolkit. Developed by Microsoft Research and founded in 2016.

一个免费且易于使用的开源和深度学习工具包 。 由Microsoft Research开发，于2016年成立。

特点： (Characteristics:)

• Written In: C++ and Python

  撰写： C ++和Python

• Platforms: Windows, Linux

  平台： Windows，Linux

• Model Types: Support feed-forwards DNNs, convolutional and recurrent neural networks

  模型类型：支持前馈DNN，卷积和递归神经网络

关键点： (Key points:)

• Efficient for image recognition高效的图像识别
• Efficient for voice recognition高效的语音识别
• Efficient for handwriting recognition高效的手写识别
• Used and known by big companies for its speech recognition, predictive analysis, and understanding large unstructured datasets, capabilities.被大公司使用并因其语音识别，预测分析以及理解大型非结构化数据集和功能而闻名。

Caffe2 (Caffe2)

Caffe2 is a flexible deep learning framework developed by Berkeley Vision and Learning Center and recently part of PyTorch.

Caffe2是由Berkeley Vision and Learning Center开发的灵活的深度学习框架，并且是PyTorch的一部分。

特点： (Characteristics:)

• Written In: C++ and Python

  撰写： C ++和Python

• Platforms: Linux, macOS, Windows

  平台： Linux，macOS，Windows

关键点： (Key points:)

• Easy to use and quick.易于使用和快速。
• Flexible and high-performance framework for your machine learning product uses.适用于您的机器学习产品的灵活而高性能的框架。
• Modular and facilitate fast prototyping of ideas and experiments in deep learning.模块化并促进深度学习中思想和实验的快速原型制作。

阿帕奇·辛加(Apache Singa) (Apache Singa)

Apache Singa is a scalable and flexible Deep Learning platform developed by Apache Software Foundation and introduced in 2015.

Apache Singa是由Apache Software Foundation开发并于2015年引入的可扩展且灵活的深度学习平台。

特点： (Characteristics:)

• Written In: C++, Python, Java

  撰写： C ++，Python，Java

• Platforms: Linux, macOS, Windows

  平台： Linux，macOS，Windows

关键点： (Key points:)

• Scalability: SINGA parallelizes the training and optimizes the communication cost to improve training scalability.

  可扩展性：SINGA 使培训并行化并优化通信成本，以提高培训的可扩展性。

• Efficiency: builds a computational graph to optimizes the training speed and memory footprint.

  效率： 建立计算图以优化训练速度和内存占用。

• Usability: has a simple software stack and Python interface to improve usability.

  可用性：具有简单的软件堆栈和Python界面以提高可用性。

外卖： (Takeouts:)

• There is no doubt that AI is transforming the IT world.毫无疑问，人工智能正在改变IT世界。
• It evident too that it will bring a lot of value and power to web development.同样很明显，它将为Web开发带来很多价值和力量。
• Stats are showing the big competition around this field.统计数据显示该领域的巨大竞争。
• Tomorrow web developers will be more competitive with AI and ML skills.明天的Web开发人员将在AI和ML技能方面更具竞争力。
• Web applications will become smarter with AI.借助AI，Web应用程序将变得更加智能。

• According to this article, 84% of businesses say that Artificial Intelligence will enable them to sustain or obtain a competitive advantage.

  根据这篇文章 ，84％的企业表示人工智能将使他们能够维持或获得竞争优势。

• According to this article, 72% of respondents in the technology, media, and telecommunications industry expect AI to have a significant impact on product offerings in the next five years.

  根据本文所述，技术，媒体和电信行业中72％的受访者预计AI在未来五年内将对产品提供产生重大影响。

Dear reader, thank you for your time. Follow me for more interesting stories about web development and AI \ (•◡•) /

亲爱的读者，谢谢您的宝贵时间。 跟随我，获得有关Web开发和AI的更多有趣故事\(•◡•)/

See you soon (ﾉ◕ヮ◕)ﾉ*:･ﾟ✧

很快见(ﾉ◕ヮ◕)ﾉ *：･ ﾟ✧

FAM

FAM