Machine Learning at AWS re:Invent 2021

 · 4 mins read

Contents


Photo by Samuel Pereira on Unsplash

Machine Learning at AWS re:Invent 2021

An overview of some of the most interesting upcoming sessions at re:Invent 2021 regarding Machine Learning & Data Science

Introduction

AWS re:Invent is each year one of the most awaited conferences by tech enthusiasts and for 2021 (29th Nov - 3rd Dec) is back with both a virtual and in person version. As part of this year event many different topics are going to be covered such as: Data, AI, Security, DevOps, etc…


Photo by Mehmet Ali Peker on Unsplash

My Favorite picks

In order to prepare best for the event, I had the opportunity to go through most of the upcoming session and picked my 5 suggestions for you! Before getting started, if you are completely new to AWS and its Data Science & Machine Learning services, I previously wrote this guide in order to kickstart your AI journey with AWS.

Use Amazon SageMaker to develop high-quality ML models faster (AIM301 - Breakout Session)

Amazon SageMaker is the AWS platform designed in order to help Data Scientists build, train, and deploy Machine Learning models. SageMaker has been proved to increase teams productivity of up to 10 times, thanks to its various built-in capabilities such as data labelling, interactive data preparation, Feature Stores, workflows, and human in the loop. As part of this session, you will have the opportunity to get an introduction to Sagemaker and it’s various capabilities.

Feel free to save the session on your calendar at this link in order to make sure to don’t miss it!

Applying AWS machine learning to next-gen DevOps (AIM312 - Chalk Talk)

DevOps has been one of the major trends in the tech world over the last few years, completely revolutionizing how software is developed and maintained. Although, problems related to security and handling of user sensitive information can still happen if testing techniques and code reviews are not carried out correctly. As part of this session, you will learn how AWS is trying to overcome these kind of problems, by developing Machine Learning augmented DevOps tools to automate these tasks and spot anomalies in the codebase.

Feel free to save the session on your calendar at this link in order to make sure to don’t miss it!

Implementing MLOps practices with Amazon SageMaker (AIM320 - Breakout Session)

MLOps (Machine Learning -> Operations) is a set of processes designed to transform experimental Machine Learning models into productionized services ready to make decisions in the real world. At his core, MLOps is based on the same principles of DevOps but with an additional focus on data validation and continuous training/evaluation. Some of the main advantages of MLOps are:

  • Focus on flexibility of training/comparing different ML models.
  • Improved time to market.
  • Increased model robustness (easier to identify data drift, retraining models, etc.).

Going through this session, you will have the opportunity to learn how Amazon SageMaker Pipelines can help you boost automation, version ML models, improve governance and keep track of your data lineage.

Feel free to save the session on your calendar at this link in order to make sure to don’t miss it!

Gain greater visibility and transparency into your ML models (AIM322 - Chalk Talk)

Application of Machine Learning in ambits such as medicine, finance and education is still nowadays quite complicated due to the ethical concerns surrounding the use of algorithms as automatic decision-making tools. Two of the main causes at the root of this mistrust are: bias and low explainability. In this session, you will have an introduction to SageMaker Clarify and how it can be used in order to identify biases in your model and which features might have greater weight on your model behavior depending on different type of inputs.

Feel free to save the session on your calendar at this link in order to make sure to don’t miss it!

Improve ML productivity through better feature engineering (AIM411 - Chalk Talk)

Feature Engineering, alongside with Features Selection and Extraction, is one of the most important steps in a Machine Learning pipeline. These steps are in fact focused on creating the best possible input data to feed into a Machine Learning model, making therefore easier to achieve good results without having necessarily to use overly complicated models. As part of this session, you will have an introduction to Amazon SageMaker Feature Store and how it can be used in order to share ML features across your organizations and reuse them for different applications.

Feel free to save the session on your calendar at this link in order to make sure to don’t miss it!

Contacts

If you want to keep updated with my latest articles and projects follow me on Medium and subscribe to my mailing list. These are some of my contacts details:

Buy Me A Coffee

Subscribe to my Data Blog Newsletter to get news about Data, Cloud and my website!

* indicates required