Machine Learning

Co-authors: Javen Xu, Hakan Baba and Adriana Deneault Intro Graph learning methods can reveal interesting insights that capture the underlying relational structures. Graph learning methods have many industry applications in areas such as product or content recommender systems and network analysis. In this post, we discuss how we use graph learning…

A few weeks ago, I remember watching a video by DJ Trev about building the foundation of a house. The foundation of a house is the most important element of the entire structure. It serves as the base upon which everything else is built, and provides stability and support for…

Leveraging early stopping for LightGBM, XGBoost, and CatBoost — Gradient-boosted decision trees (GBDTs) currently outperform deep learning in tabular-data problems, with popular implementations such as LightGBM, XGBoost, and CatBoost dominating Kaggle competitions [1]. Early stopping — a popular technique in deep learning — can also be used when training and tuning GBDTs. However, it is common to see practitioners…

Airbnb Categories Blog Series — Part II : ML Categorization By: Mihajlo Grbovic, Pei Xiong, Pratiksha Kadam, Ying Xiao, Aaron Yin, Weiping Peng, Shukun Yang, Chen Qian, Haowei Zhang, Sebastien Dubois, Nate Ney, James Furnary, Mark Giangreco, Nate Rosenthal, Cole Baker, Bill Ulammandakh, Shankar Shetty, Sid Reddy, Egor Pakhomov Airbnb…

Some ChatGPT users have encountered issues such as disappearing dialogue history, while others have faced a different problem: seeing the dialogue topics of other users (it’s advisable not to discuss the private key of your crypto wallet with ChatGPT). To solve these issues we should interact with GPT via API…

What works, the pros and cons, and example code for each approach — If some of the terminology I use here is unfamiliar, I encourage you to read my earlier article on LLMs first. There are teams that are employing ChatGPT or its competitors (Anthropic, Google’s Flan T5 or PaLM, Meta’s LLaMA, Cohere, AI21Labs, etc.) for real rather for cutesy demos. Unfortunately, informative…

There’s a lot of justifiable excitement around using LLMs as the basis for building a new generation of smart apps. While tools like LangChain or GPT-Index makes it easy to create prototypes, it’s useful to look behind the curtain to see what the wizard looks like before taking these apps…