Implementing custom models for machine learning
summary
Announces MemSQL 6 with three concrete changes: enhanced online operations that keep mission‑critical data available, vectorized execution and compact query processing that speed analytics, and an extensibility layer with stored procedures so users implement custom models (examples: dot product for native vector ops and k‑means in MemSQL syntax). Shows a Spark connector workflow: train or stream models in Spark, push results into MemSQL for low‑latency scoring and many concurrent clients.
AI-written summary · about
chapters (6)
- 0:00Introduction
- 1:03Machine learning
- 1:48Realtime
- 2:27Secret sauce
- 3:31Industry use cases
- 4:33Whats next for MemSQL
related talks
05mDriving the On-Demand Economy with Predictive Analytics
09mHow to be fast and have a good memory with Machine Learning
43mNatively Engineered for AI: The Evolution
06mMachines and the magic of fast learning
12mOpenMLDB: An Open-Source Real-Time Feature Platform
21mWhat If We Could Use Machine Learning Models as Tables?
39mMachine Learning with Scala on Spark
51mDiscover BigQuery ML, build your own CREATE MODEL stmt