Getting started with machine learning in Postgres using Hasura and PostgresML смотреть онлайн

In this video:
00:00 Recap
00:50 Welcome
02:15 David's Background
05:00 When to choose which type of database
08:50 Why relational data structures are still dominant
16:45 Relational data with graph-like queries
18:00 Analytical workloads on relational data structures
21:00 BI, Analytics, and Statistical models
27:10 Where to write my code
29:50 OLTP vs OLAP, data lakes vs data warehouses
34:30 The frontier of ML moving to built-in functionality
38:30 Standardizing SQL flavors with Hasura
40:00 Introduction to PostgresML
43:00 Looking at our datasets
48:00 Writing probabilistic functions in PgML
53:00 Volatile vs Stable vs Pure functions
00:55 Adding the functions in Hasura
1:00:00 Closing

Jesse and David talk about the history of relational databases, the shift of machine learning functionality to being built in on the database, and write some probabilistic functions that get automatically exposed through our Hasura API.

This is a great video for anyone looking to see what the data-science puck is moving in the world of databases and shows how Hasura unlocks your database's secret powers with very little effort and code.


Links:
https://postgresml.org/
https://www.nltk.org/
https://scikit-learn.org/stable/
https://greenplum.org/
https://aws.amazon.com/redshift/

-----

? Hasura makes your data instantly accessible over a real-time GraphQL API, so you can build and ship modern apps and APIs faster. Hasura connects to your databases, REST servers, GraphQL servers and third party APIs (eg: Stripe, Salesforce) to provide a unified realtime GraphQL API across all your data sources instantly.

Check out our features: https://hasura.io
Find us on Twitter: https://twitter.com/HasuraHQ
Join our community: https://hasura.io/discord

-----