Shulou(Shulou.com)06/03 Report--

Machine Learning in SAP Strategy

Https://mp.weixin.qq.com/s/buv3R6AEr5PWhjpniYh_BQ

The author has introduced why SAP brings the vision of smart enterprises to market and helps enterprises meet the challenges of digitization.

The purpose of this blog is to try to explain the basic concepts of machine learning and how to use it in SAP's portfolio.

A key element of intelligent enterprises is to inject machine learning algorithms into enterprise software business processes. Machine learning algorithms themselves are nothing new, but they have been in a very hot state recently because we are in a stage where technology is advancing by leaps and bounds and a large number of applications can be landed.

In the past 30 or 50 years, enterprises have produced a lot of data, and in the last 10 years, the production of enterprise data is getting faster and faster. These data give birth to the dawn of the digital economy. If they are trained and handled effectively, they will be rewarded handsomely. Bring value to the enterprise.

1.

What is machine learning?

The goal of the machine learning algorithm is to determine a mathematical model, which we call f (), in y = f (x), where x represents any observation of the real world, and y represents some evaluation, identification or processing by those x.

There are three main types of machine learning algorithms: supervised learning, semi-supervised learning and unsupervised learning.

Supervised learning works with tagged data, which means that by analyzing the training data, the equation y = f (x) based on real-world observation (x) is found to predict / infer certain results (y).

Semi-supervised learning works with some labeled data and many unlabeled data, which means finding the equation y = f (x), even if the resulting value (y) is known only by a limited set of observations (x), such as in a sampling plan.

Unsupervised learning only uses (x) input data, which means finding hidden structures or observing relationships (x). The variable (y) is automatically generated from the analysis.

To add another layer of complexity, the function f () we intend to define is usually affected by many variables and may have dependencies on each other: y = f (x1m x2m x3, … , xn.

The machine learning algorithm establishes the mathematical model f () according to the training data, and the ultimate goal is to make statistical prediction or decision without clear programming.

The "training" model then means determining the "correct" values (y) of all (xn) from the example of the training data set. For example, in supervised learning, machine learning algorithms build models by examining many labeled data (xn) and trying to find a model that minimizes loss.

Losses are numbers that indicate how bad the model predicts on a single example (a bad (y)). If the prediction of the model is perfect, then the loss will be zero; otherwise, the loss will be high. The purpose of the training model is to find a set of low loss (xn), and then infer the correct (y).

The following figure shows the mechanism for training machine learning models:

This is why the larger the training data set, the better the data quality, so that we can more accurately collect the basic conditions of the function f () and improve the accuracy of prediction.

Therefore, in order to make machine learning successful, we need to prepare a large amount of high-quality data.

two。

Machine Learning of SAP

SAP's machine learning vision includes:

Intelligent business processes embedded in ERP Sramp 4 HANA and LOB SaaS tools (SuccessFactors,Ariba,C / 4HANA MagneConcur, etc.)

Intelligent digital platform, which is powered by chat robots, virtual assistants, robot process automation, and provides machine learning algorithms developed by SAP, and through our partners such as Google's TensorFlow is a huge library.

Finally, SAP HANA serves as the basis for preparing, managing, ensuring the quality of all data governance engines required, and real-time processing the data needed to train your own predictive analysis engine and other machine learning algorithms can be obtained from SAP Cloud Platform.

To summarize this overview, I'll introduce SAP's main products around machine learning.

3.

SAP Leonardo Machine Learning Foundation

SAP Leonardo Machine Learning Foundation is a SAP Cloud Platform-based machine learning platform that implements simple functions and provides tight integration with the SAP backend. Its four main functions are:

Easy to use ML content for non-ML experts, allowing you to deploy and run your own machine learning model, or use your own data to adapt existing models.

Image processing service to enable automatic mode detection

A text processing service that analyzes natural language content stored in a document, website or email and reveals its meaning.

Voice processing service that converts voice to text and combines it into a digital assistant or voice control application for use.

4.

Process Automation of SAP Intelligent Robot

SAP Intelligent process Automation (IPA) is also based on the SAP cloud platform, integrating robot process automation and machine learning into an integrated automation product:

ML "thinks" and processes unstructured processes, data, and improves a pure rule-based decision engine.

RPA "Action", especially the execution of business processes across systems, allows key users to create their own robots.

The purpose of IPA is to accelerate the digital transformation of business processes by automatically copying tedious behaviors with no added value. It involves business process automation across applications and systems.

5.

SAP Conversational AI

SAP Conversational AI is a SAP Cloud Platform-based platform for developing, deploying and monitoring Conversational AI applications. It is equipped with powerful natural language processing (NLP) technology, so you can quickly and easily build robots that really understand people.

The robot is proficient in English, French, Spanish and German and provides standard functions in 15 other languages.

Last

The use of machine learning to improve and automate business processes is limited only by our imagination. This is ultimately what smart business is all about. In addition to the actual technology, the goal of all these tools is to help you and create value for your business.

-End-

Original address: https://blogs.sap.com/2019/04/05/machine-learning-in-sap-strategy/

Original author: Arnaud SERGENT

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.