Sunday, February 20, 2022

AutoML and Production ML (MLOps) - Increase Effectiveness and Productivity of Machine Learning Projects


AutoML


AutoML platforms push data science projects to the finish line

Data science projects often have trouble reaching the production phase, but automated machine learning platforms are accelerating data scientists' work to help them come to fruition.

By Cameron Hashemi-Pour, 16 Feb 2022

AutoML speeds up the work data scientists perform through automation. Azure's AutoML is one such product offering. The automated ML software expedites and simplifies this otherwise arduous work.  

According to Dennis Michael Sawyers, author of a book on AutoML,  automated machine learning as a concept will become a trending topic in 2022 due to the data science labor shortage. For existing data scientists,  automating  much of their labor as possible will increase their productivity. AutoML lets data scientists build models very quickly and also lets new data scientists on-ramp very quickly. 

Databricks AutoML, DataRobot and Google Vertex AI AutoML are the biggest competitors to Azure AutoML.

Databricks'  AutoML feature  will surely take off due to Databricks' large and established user base. GCP positions AI as its main strength and has fairly advanced AutoML capabilities across multiple categories of data, including tabular, video, text and images. DataRobot is the most popular AutoML vendor and is mostly focused on making machine learning as accessible as possible to companies even if they lack data scientists. 

Most machine learning problems are being handled fairly well by AutoML these days. It makes sense that  data scientists use AutoML before building out custom models. For experienced data scientists,  learning curve to using AutoML isn't very steep. 

Dennis Michael Sawyers, Author, 'Automated Machine Learning with Microsoft Azure'

https://www.techtarget.com/searchenterpriseai/feature/AutoML-platforms-push-data-science-projects-to-the-finish-line

Google AutoML

Train high-quality custom machine learning models with minimal effort and machine learning expertise.

Train custom machine learning models

AutoML enables developers with limited machine learning expertise to train high-quality models specific to their business needs. Build your own custom machine learning model in minutes.

Vertex AI

Unified platform to help you build, deploy and scale more AI models.

AutoML Image

Derive insights from object detection and image classification, in the cloud or at the edge. 

 AutoML Video

Enable powerful content discovery and engaging video experiences. 

AutoML Text

Reveal the structure and meaning of text through machine learning. 

AutoML Translation

Dynamically detect and translate between languages. 

AutoML Tabular

Automatically build and deploy state-of-the-art machine learning models on structured data. 


What is AutoML?

Automated Machine Learning provides methods and processes to make Machine Learning available for non-Machine Learning experts and  to improve efficiency (productivity) of Machine Learning

Machine learning (ML) has achieved considerable successes in recent years in variety of disciplines.  Machine learning experts to perform the following tasks:

Preprocess and clean the data.

Select and construct appropriate features.

Select an appropriate model family.

Optimize model hyperparameters.

Design the topology of neural networks if they are used. 

Postprocess machine learning models.

Critically analyze the results obtained.

The rapid growth of machine learning applications has created a demand for off-the-shelf machine learning methods that can be used easily with less knowledge and effort. The resulting research area  targets progressive automation of machine learning and it is termed AutoML.

See for some outputs of AutoML Research and some presentations

AutoML - Conferences


AutoML-Conf 2022
1st International Conference on Automated Machine Learning
 Baltimore, US (co-located with ICML; we're currently planning an in-person event)
 July 25, 2022 - July 27, 2022

Production ML - MLOps


Production ML: Getting Started with MLOps

How Can MLOps help you productionize your ML projects? And Why do you need to start adopting MLOps ASAP?

Hajar Khizou, Feb 17, 2022

MLOps tends to be defined as the counterpart of DevOps for machine learning. We can therefore deduce that MLOps allows the automation and monitoring of the steps of a machine learning project.

MLOps is the operationalization of Machine Learning model management. It aims to create an end-to-end process for creating, implementing, and managing repeatable, testable, and scalable machine learning models.


















No comments:

Post a Comment