Flender

Opening the “bag of tricks” that makes a modern recommender system work(Logo design Eyestetix Studio, background design by Dan Cristian Pădureț)Recommender systems have become one of the most ubiquitous industrial Machine Learning applications of our times, but little is being published about how they actually work in practice.A notable exception is Paul Covington’s paper “Deep Neural Networks for YouTube Recommendations”, which is packed with numerous practical insights and learnings about YouTube’s deep-learning powered…

The quest to build Machine Learning systems that learn like us(Midjourney)The “holy grail” of Machine Learning is the ability to build systems that can both memorize known patterns in the training data as well as generalize to unknown patterns in the wild.It’s the holy grail because this is how we humans learn as well. You can recognize your grandma in an old photo, but you could also recognize a Xoloitzcuintli as a dog even though you’ve never actually seen one before. Without memorization we’d have to constantly…

Understanding the most underrated trick in applied Machine Learning(Midjourney)Hashing is one of the most common “tricks” used in industrial Machine Learning applications, yet it doesn’t get nearly as much attention as it deserves.The biggest advantage of hashing, especially in modern recommender systems, is its finite-memory guarantee: without hashing, it would be highly impractical to learn the relevance of billions of videos, news articles, photos, or web pages for billions of users without running out of memory.But…

Behind the ongoing quest for building unbiased models from biased dataImage generated by the author with MidjourneyRecommender systems have become ubiquitous in our daily lives, from online shopping to social media to entertainment platforms. These systems use complex algorithms to analyze historic user engagement data and make recommendations based on their inferred preferences and behaviors.While these systems can be incredibly useful in helping users discover new content or products, they are not without their flaws:…

A practical guide on how to optimize your experiments for business impactImage source: PexelsExperimentation is at the heart of the Machine Learning profession. We progress because we experiment.However, not all experiments are equally meaningful. Some create more business impact than others. Yet, the art of selecting, executing, and iterating on experiments with a focus on impact isn’t usually covered in standard ML curricula.This creates a lot of confusion. New ML practitioners may get the impression that you’re…

A primer on position bias (and why it matters)Image generated with Stable DiffusionStandard Machine Learning curricula teach that ML models learn from patterns that exist in the past in order make predictions about the future.This is a neat simplification, but things change dramatically once the predictions from these models are being used in production, where they create feedback loops: now, the model predictions themselves are impacting the world that the model is trying to learn from. Our models no longer just predict…

Questions you should ask and red flags you should avoidPhoto by Adi Goldstein on UnsplashAs Machine Learning professional, navigating the diverse landscape of ML roles within the industry can be confusing. Job titles are usually not a big help because they change depending on the company and also depending on the organization within a company. Job titles tend to change over time as well, as we’ve seen in the rebranding of data analysts to data scientists.In order to navigate the job market and find potential roles for…

Eliminate your recurring pain points by understanding the underlying patternsPhoto by Jeremy Bishop on UnsplashAnti-patterns are common yet counter-productive responses to recurring problems. Because they’re ineffective, they perpetuate recurring pain points without ever resolving the underlying, systematic issues. Anti-patterns exist exist pretty much anywhere people come together to solve problems, in software development, project management, and yes, also in data science.Knowledge about anti-patterns is the best way to…

Behind the algorithms that make Machine Learning models bigger, better, and fasterImage generated with Stable DiffusionDistributed learning is one of the most critical components in the ML stack of modern tech companies: by parallelizing over a large number of machines, one can train bigger models on more data faster, unlocking higher-quality production models with more rapid iteration cycles.But don’t just take my word for it. Take Twitter’s:Using customized distributed training allows us to iterate faster and train…

The most useful computer science class you’ve probably never takenImage generated with Stable DiffusionOne thing that I have consistently observed throughout my career is that the most productive data scientists and engineers have usually one thing in common: they’re command-line wizards. They can navigate their computer’s file system, search for patterns in log files, and manage jobs, source code, and version control all from the command line, without relying on slow navigation with the mouse and graphical user…