How to Build, Train and Deploy Your Own Recommender System â Part 2
We build a recommender system from the ground up with matrix factorization for implicit feedback systems. We then deploy the model to production in AWS.
This is my Day 1 report for the AWS Summit which is an all Online event this year.
Even if the event was a real face-to-face conference, I would still have been attending the sessions online, since coincidentally, Iâve very recently contracted Covid, and I am still recovering as Iâm typing this report.
I make no secret that my goal in the next 12 months is to build up Machine Learning skills and find projects that will exploit these newfound skills. So many of my learning activities this past year have been about this field, and the effort is still continuing. Imagine my surprise when I found out that the AWS Summit has a good chunk of sessions focusing solely on AI/ML!
Out of about 104 sessions, 28 are about AI/ML, I thought that was great, so this for yearâs AWS Summit, thatâs where my focus will be. Obviously I could not have attended all 28 sessions, just the ones that I fancied and handpicked from the agenda.
The first session is a peek into how Canva approached scaling ML in their company. As a newcomer in the filed of ML, it was
This can be summarized into 5 main points, please see image below. When the AWS Summit videos are made available, I will post their links here.
Team Enablement - involves defining culture, team training (and hackathon), and tools
Data Strategy - involves training, best (data) practice
Proof of Concept - involves training, business alignment, cross functional teams
Repeatability - involves training, tools, cloud enablement, MLOps
Scale - involves training, Center of Excellence, optimization
AWS and Intel have a deep partnership and together provide solutions to real world customer problems. Many of the these are related to AI and ML, such as in the development of hardware chips that provide faster, more powerful performance that is cheaper and more sustainable to run.
This session introduces the concept of Machine Learning Operations (MLOps) in Data Science/Machine Learning teams. As the teamâs AI maturity improves, it becomes more and more important to integrate MLOps practices into your workflow to avoid the problems of POCs not making it into production.
Issues like manual intervention for changes, work stuck in Jupyter notebooks, not enabling autoscaling, or having no feedback on model performance are only some of the issues that contribute to the failure of your ML system.
We were then presented with 3 MLOps architectures, from Small, Medium and Large, which represents architectures that one would adopt, depending on your teamâs AI maturity, and budget.
In re:Invent 2021, AWS announced that Sustainability as the latest pillar added to the 5 existing: Operational Excellence, Security, Reliability, Performance Efficiency and Cost Optimization. It was nice to see sustainability being more and more involved in presentations, so it was one that I was looking forward to hear.
Leadership is crucial if sustainability initiatives are to succeed. To adopt a sustainability mindset, direction from leaders is a requirement. From there the company can enable teams, adopt a powerful data platform such as AWS, and finally choosing the right first project.
The session also showed an example in an industrial setting of one Amazon.comâs buildings. The problem was to improve the current systemâs power consumption by 8% and achieved that by creating a model based on Reinforcement learning and Amazon SageMaker suite of tools.
Not everyday does one have the chance to develop systems used in autonomous vehicles and AWS have just given every developer the ability to do so in DeepRacer! Using Amazon SageMaker and Intelâs OpenVINO toolkit, Reinforcement learning can be used to train your models, after which it can be run in the DeepRacer 3D racing simulator.
Also, AWS has purpose-built a real 1/18th scale race car, so that you can use it to download your models to, and race against the autonomous racing league.
Thatâs a wrap for Day 1 of the AWS Summit 2022 Australia and New Zealand. Iâve got two highlights for today.
First is seeing how AWS SageMaker is being positioned by Amazon at the center of their AI/Machine Learning story - as an end to end MLOps framework. When I learned about Data Science in a course I completed towards the end of 2021, I was not made aware of such tool available, and rightly so. It would have made the course more difficult than it was.
And next was seeing DeepRacer - truly making Machine Learning (Reinforcement Learning) accessible to developers the world over.
Keep an eye out for my Day 2 wrap later!
We build a recommender system from the ground up with matrix factorization for implicit feedback systems. We then deploy the model to production in AWS.
We build a recommender system from the ground up with matrix factorization for implicit feedback systems. We put it all together with Metaflow and used Comet...
Building and maintaining a recommender system that is tuned to your businessâ products or services can take great effort. The good news is that AWS can do th...
Provided in 6 weekly installments, we will cover current and relevant topics relating to ethics in data
Get your ML application to production quicker with Amazon Rekognition and AWS Amplify
(Re)Learning how to create conceptual models when building software
A scalable (and cost-effective) strategy to transition your Machine Learning project from prototype to production
An Approach to Effective and Scalable MLOps when youâre not a Giant like Google
Day 2 summary - AI/ML edition
Day 1 summary - AI/ML edition
What is Module Federation and why itâs perfect for building your Micro-frontend project
What you always wanted to know about Monorepos but were too afraid to ask
Using Github Actions as a practical (and Free*) MLOps Workflow tool for your Data Pipeline. This completes the Data Science Bootcamp Series
Final week of the General Assembly Data Science bootcamp, and the Capstone Project has been completed!
Fifth and Sixth week, and we are now working with Machine Learning algorithms and a Capstone Project update
Fourth week into the GA Data Science bootcamp, and we find out why we have to do data visualizations at all
On the third week of the GA Data Science bootcamp, we explore ideas for the Capstone Project
We explore Exploratory Data Analysis in Pandas and start thinking about the course Capstone Project
Follow along as I go through General Assemblyâs 10-week Data Science Bootcamp
Updating Context will re-render context consumers, only in this example, it doesnât
Static Site Generation, Server Side Render or Client Side Render, whatâs the difference?
How to ace your Core Web Vitals without breaking the bank, hint, its FREE! With Netlify, Github and GatsbyJS.
Follow along as I implement DynamoDB Single-Table Design - find out the tools and methods I use to make the process easier, and finally the light-bulb moment...
Use DynamoDB as it was intended, now!
A GraphQL web client in ReactJS and Apollo
From source to cloud using Serverless and Github Actions
How GraphQL promotes thoughtful software development practices
Why you might not need external state management libraries anymore
My thoughts on the AWS Certified Developer - Associate Exam, is it worth the effort?
Running Lighthouse on this blog to identify opportunities for improvement
Use the power of influence to move people even without a title
Real world case studies on effects of improving website performance
Speeding up your site is easy if you know what to focus on. Follow along as I explore the performance optimization maze, and find 3 awesome tips inside (plus...
Tools for identifying performance gaps and formulating your performance budget
Why web performance matters and what that means to your bottom line
How to easily clear your Redis cache remotely from a Windows machine with Powershell
Trials with Docker and Umbraco for building a portable development environment, plus find 4 handy tips inside!
How to create a low cost, highly available CDN solution for your image handling needs in no time at all.
What is the BFF pattern and why you need it.