Jun Yu

Jun Yu

Engineering Leader @Whatnot | ex-{Snap, Amazon, eBay}

Sitemap

A list of all the posts and pages found on the site. For you robots out there is an XML version available for digesting as well.

Pages

Posts

Growth = Step-Function Changes + Incremental Improvements

6 minute read

Published:

October 2024 marks my fifth year at Snap, making this my second-longest tenure with an organization, aside from my graduate program. While there are still many challenges to tackle at work, I want to take a moment to reflect on the journey that brought me here and to share some of the insights and lessons I’ve gathered along the way—many of which were learned the hard way. In this post, I will share my observations on growth.

Hello World

less than 1 minute read

Published:

Hi, I started this blog with the goal of recording a few of the lessons I’ve learned in my life. They ranges from learnings from work and parenting, thoughts from reading a book or post, or oftentimes just some random thoughts. I use the term “Gradient” to describe a lession which may sound nerdy, but I find learning life lessions is similar to how gradient descent works in machine learning. They both involve making changes to the path in seeking the best outcome (or the global optimal). Plus, I guess I am nerdy inside. If you are nerdy and feel bad about it, you should check out Paul Graham’s post on this topic. It will for sure change your perspective. Lastly, these posts are really just for myself in recording some of my learnings in seeing the world. This way I can be truthful with myself and do not have to fight the desire to make myself look better.

portfolio

publications

Embedding Based Retrieval in Friend Recommendation.

Published in Proceedings of the 46th International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR), 2023

Jiahui Shi, Vivek Chaurasiya, Yozen Liu, Shubham Vij, Yan Wu, Satya Kanduri, Neil Shab, Peicheng Yu, Nik Srivastava, Lei Shi, Ganesh Venkataraman, and Jun Yu

talks

Latent Dirichlet Allocation-based Diversified Retrieval for E-commerce Search.

Published:

Diversified retrieval is a very important problem on many e-commerce sites, e.g. eBay and Amazon. Using IR approaches without optimizing for diversity results in a clutter of redundant items that belong to the same products. Most existing product taxonomies are often too noisy, with overlapping structures and non-uniform granularity, to be used directly in diversified retrieval. To address this problem, we propose a Latent Dirichlet Allocation (LDA) based diversified retrieval approach that selects diverse items based on the hidden user intents.

Bayesian and empirical Bayesian forests.

Published:

We derive ensembles of decision trees through a nonparametric Bayesian model, allowing us to view random forests as samples from a posterior distribution. This insight provides large gains in interpretability, and motivates a class of Bayesian forest (BF) algorithms that yield small but reliable performance gains. Based on the BF framework, we are able to show that high-level tree hierarchy is stable in large samples. This leads to an empirical Bayesian forest (EBF) algorithm for building approximate BFs on massive distributed datasets and we show that EBFs outperform subsampling based alternatives by a large margin.

teaching

Teaching experience 1

Undergraduate course, University 1, Department, 2014

This is a description of a teaching experience. You can use markdown like any other post.

Teaching experience 2

Workshop, University 1, Department, 2015

This is a description of a teaching experience. You can use markdown like any other post.