The name of the base system (platform): | Microsoft Azure |
Developers: | Microsoft |
Date of the premiere of the system: | 2014/07/12 |
Last Release Date: | May 2019 |
Technology: | BI, CRM |
Content |
2021: Seamq Advanced Analytics Integration
On November 10, 2021, the company Seeq announced the launch of additional integrated support for algorithms machine learning Microsoft Azure Machine Learning. More. here
2019: Update that allows you to create MO programs without programming knowledge
In early May 2019, Microsoft updated Azure Machine Learning, as a result of which the service began to provide the ability to simply create machine learning algorithms - now you don't even need to write code to develop smart programs.
There are people who study machine learning concepts and want to create their own models, but they are not programmers. It can be IT professionals or statisticians or mathematicians. For such customers, we offer the ability to visually create models, "said Bharat Sandhu, head of the artificial intelligence department at Microsoft. |
Microsoft addresses Azure Machine Learning to the following user categories:
- developers and scientists with programming skills;
- businessmen and people with knowledge of data, but unable to write code;
- users just beginning to learn the concept of machine learning.
The graphical interface implemented in Azure Machine Learning allows you to create and deploy machine learning models even without writing code and use the drag and drop function.
To create a machine learning model, you only need to load a dataset and specify what value to predict. The service automatically launches many algorithms and optimizations, but if necessary, everything can be configured manually.
Azure Machine Learning also received the following innovations:
- New capabilities of MLOps development practices (machine learning DevOps). Integration with Azure DevOps tools ensures the reproducibility, verifiability, and automation of the entire machine learning development lifecycle.
- Wide availability of hardware-accelerated models running on, FPGA (Field Programmable Gate Array) as well as support for ONNX the Runtime engine on the Nvidia TensorRT and for Intel nGraph fast interference (pin generation) on Nvidia and Intel chipsets. [1]
2014: Launch of public beta
On June 18, 2014, Microsoft announced plans to launch a publicly available beta version of Azure Machine Learning, in July 2014.
The purpose of the service is to make machine learning technologies available to a large number of users, the corporation explained. The system is able to recognize trends in a large amount of continuously arriving information, predicting their further development and future events.
The service is equipped with an interface called Machine Learning Studio, which offers typical machine learning algorithms in the form of icons from which you can compose workflows. You can also write your own algorithms in R. In addition, the "studio" allows you to create and train analytical models. Ready-made models can be published as Azure services and connected to data sources using programming interfaces. Provided SDK for third-party developers, which allows you to create arbitrary services and applications based on Azure ML.
Microsoft also reported on the previously closed beta testing of the service in a large retail trading company.
Machine Learning Capabilities for Banks:
- Predicting problems and solutions for customers' problems.
- Offering the most suitable products.
- Model risk.
- Identification of frode.