InfoQ Homepage Data Model Content on InfoQ
-
Scaling & Optimizing the Training of Predictive Models
Nicholas Mitchell presents the core building blocks of an entire toolchain able to deal with challenges of large amounts of data in an industrial scalable system.
-
Accuracy as a Failure
V. Warmerdam talks about cautionary tales of mistakes that might happen when we let data scientists on a goose chase for accuracy. Highly accurate models are more damaging than the inaccurate ones
-
Implementing AutoML Techniques at Salesforce Scale
Matthew Tovbin shows how to build ML models using AutoML (Salesforce), including techniques for automatic data processing, feature generation, model selection, hyperparameter tuning and evaluation.
-
Tools to Put Deep Learning Models in Production
Sahil Dua discusses how Booking.com supports data scientists by making it easy to put their models in production, and how they optimize their model prediction infrastructure for latency or throughput.
-
Using Bayesian Optimization to Tune Machine Learning Models
Scott Clark introduces Bayesian Global Optimization as an efficient way to optimize ML model parameters, explaining the underlying techniques and comparing it to other standard methods.
-
Data Structures in and on IPFS
Juan Batiz-Benet makes a short intro of IPFS (the InterPlanetary File System) and discusses the IPLD data model and example data structures (unixfs, keychain, post).
-
Towards Immutable Resources
Mark Derricutt discusses the importance of having different read and write data models when working with RESTful web APIs.
-
The Functional Programming Concepts in Facebook's Mobile Apps
Adam Ernst shows how his team at Facebook encountered spiraling complexities and declining reliability and decided to make the shift to functional, in the data model and the view layer of News Feed.
-
Data Modeling for Scale with Riak Data Types
Sponsored by Basho. Sean Cribbs discusses the theory behind several rich data types introduced with Riak 2.0 and then walking through some example applications that use them in popular languages.
-
Tracking Millions of Ganks in Near Real Time
Garrett Eardley explores how Riot Games is using Riak for their stats system, discussing why they chose Riak, the data model and indexes, and strategies for working with eventually consistent data.
-
How Draw Something Scaled To 50 million New Users, in 50 Days, with Zero Downtime
Robin Johnson discusses using a data management model for games that can be scaled, and the bottlenecks and challenges met by OMGPOP scaling to millions of users.
-
The Big Data Revolution
Claudia Perlich keynotes on M6D’s approach to Big Data, using data granularity to build predictive models used for user targeting, bid optimization and fraud detection.