قالب وردپرس درنا توس
Home / Business / Microsoft launches a drag-and-drop machine learning tool – TechCrunch

Microsoft launches a drag-and-drop machine learning tool – TechCrunch



Microsoft today announced three new services that all aim to simplify the process of machine learning. These range from a new interface for a tool that completely automates the process of creating models, to a new, non-code visual interface for model building, training, and deployment, all to host Jupyter advanced books for advanced users.

Getting started with machine learning is difficult. Even running the most basic of experiments takes a good amount of expertise. All these new tools greatly simplify this process by hiding the code or giving those who want to write their own code a pre-configured platform to do so.

The new interface of Azure's automated machine learning tools makes it possible to create a model as simple as importing a dataset and then telling the service what value to predict. Users do not need to write a single line of code while in backend, this updated version now supports a number of new algorithms and optimizations that should result in more accurate models. While most of this is automated, Microsoft emphasizes that the service "provides complete transparency in algorithms so that developers and computer lovers can manually override and control the process."

For those who want a little more control from go-go, Microsoft today also launched a visual interface for its Azure Machine Learning service that allows developers to build, train, and distribute machine learning models without having to touch anyone code.

This tool, the Azure Machine Learning Visual Interface, looks suspiciously like the existing Azure ML Studio, Microsoft's first staff to build a visual visual machine learning tool. In fact, the two services look the same. The company has never created this service, and almost seemed to have forgotten it, despite the fact that it always seemed a very useful tool for getting started on machine learning.

Microsoft says this new version combines the best of Azure ML Studio with Azure Machine Learning service. In practice, this means that while the interface is almost identical, the Azure Machine Learning visual interface expands what was possible with ML Studio by running at the top of the Azure Machine Learning service and adding that security, deployment and life-cycle management services.

The service provides a simple interface to clean up your data, exercise models using various algorithms, evaluate them, and finally put them into production.

Although these first two services are clear target novices, the new hosted notebooks in Azure Machine Learning are clearly aimed at the more experienced machine learning practitioner. The notebooks are delivered pre-packaged with support for the Azure Machine Learning Python SDK and run in what the company describes as a "safe, business-ready environment." While using these notebooks, it is neither trivial nor, this new feature allows developers to quickly get started without having to set up a new development environment with all the necessary cloud wells.


Source link