Next Meetup

Jesse Steinweg-Woods @ ZEFR
Jesse is a Data Scientist at Tronc. In this meeting he will discuss using machine learning for news recommendations. 6:30 PM Doors open 6:30 PM - 7:00 PM Snack and drinks served 7:00 PM - 8:00 PM Presentation and Q&A 8:00 PM - 8:30 PM Mingling 8:30 PM Doors close Note: This event is being held both at a different location and different time than previous events. I hope the new time will allow more people to attend. Title: News Recommendations at Tronc with Online Learning Abstract: Recommender systems are present in a variety of industries with the explicit goal of increasing a user's interaction with a website or consumption of a product. For news recommendations, constructing an effective recommender system can be a challenging problem due to a rapidly changing catalog of available items to recommend, most of which have a very short time period of being interesting or relevant to the user. Unlike the usual approach to recommender systems involving collaborative filtering via matrix factorization, we instead treat recommendations as a sequential extreme multi-class classification problem. By designing a news recommender system as a classification problem, this provides a more effective way of recommending news articles that allows flexibility and scale. It also overcomes the traditional issues found in collaborative filtering, which is especially helpful in smaller news markets. In addition, the rapid update frequency of the system allows it to adjust quickly in breaking news situations. In this talk, I will discuss the infrastructure and modeling strategy of the news recommender system currently used at Tronc along with lessons learned from scaling and serving recommendations to millions of users a day. Bio: Jesse Steinweg-Woods is a Senior Data Scientist at Tronc, a media company that includes the Los Angeles Times along with 9 other papers and a variety of websites. He works on building both internal and consumer-facing products utilizing machine learning. Areas of focus include recommender systems, article popularity prediction, unsupervised learning applied to content, and churn modeling of subscribers. He received his Ph.D. in Atmospheric Science from Texas A&M University, where he did research on numerical weather and climate prediction. Parking and directions: There is street parking in the neighborhood east of Redwood Ave. Please enter through the doors at 4104 Redwood Ave.


4101 Redwood Ave · Marina Del Rey, ca

    Past Meetups (129)

    What we're about

    Public Group

    The LA Machine Learning Meetup began with events discussing a large variety of machine learning topics such as classification, clustering, neural networks, graphical algorithms, information retrieval, search, game theory, computational learning theory, reinforcement learning, collaborative filtering etc. More recently, a Data Science Track ( was launched with the focus on business applications and the entire process of data mining (e.g. business understanding, data collection, exploratory data analysis, data transformations, feature engineering, modeling, model validation, deployment, communication of results). While in fact Machine Learning is a part of Data Science and not the other way around, rather than starting a new meetup group, Data Science is featured within the already existing Machine Learning meetup.

    Members (2,903)

    Photos (175)