Brief information about Azure Machine Learning

What is Azure Machine Learning?

A cloud-based environment You can use to train, deploy, automate, manage and track machine learning models. There are 2 options for building a machine learning model:

  • Use the graphical interface with low-code and no-code experience
  • Implement by Python or R with the SDKs

What is Azure Machine Learning studio?

A web portal in Azure Machine Learning that contains low-code and no-code options for project authoring and asset management

Some components in the studio

Notebooks:

Write and run your code in Jupyter Notebook servers that are directly integrated in the studio

Azure Machine Learning designer:

Provide a graphical interface. You can drag and drop datasets and modules to create machine learning pipelines

Automated machine learning UI:

Help you choose the best model by iterating over many combinations of algorithms and hyper-parameters, based on a success metric

Data labeling:

Used for creating, managing, monitoring labeling projects such as image classification (includes multi-label and multi-class), object identification with bounded boxes

Data stores:

A reference to the data source location along with a copy of its metadata

Datasets:

Used by Data stores to securely connect to Azure storage services.

Environments:

Encapsulate the environment where training or scoring of your machine learning model happens. It specifies the Python packages, environment variables and software settings.

Runs:

A run is a single execution of a training script. The following information will be stored:

  • Metadata about the run: timestamp, duration
  • Metrics that are logged by your script
  • Output files that are auto collected by the experiment or explicitly uploaded by you
  • A snapshot of the directory that contains your scripts, prior to the run

Experiments:

A experiment is a grouping of many runs from a specified script

Models:

Include machine learning models. A model is a piece of code that takes an input and produces output

Compute resources:

A machine or set of machines you use to run your training script or host your service deployment. You can create and configure your own compute resources.

Pipelines:

Help you to create and manage workflows that stitch together machine learning phases. A pipeline might include data preparation, model training, model deployment and inference/scoring phases