Even more crazy is that {modeldb} can fit a multiple linear regression and K-means inside the database.
https://tidymodels.github.io/modeldb/index.html …
@theotheredgar is behind all of this.
-
-
- Još 1 odgovor
Novi razgovor -
-
-
In practice exporting the data and scoring in parallel is more tractable, depending on your db arch and typical load. A SQL string with low precision tree ensemble can be 100s of megabytes. Tried a few strategies for boosting. Caused a few db reboots.
-
Again, depends on the db and the UDF story what's really possible. We got further with some distributed DBs but in the end brought the data to the code. I'd be interested in hearing of a db *designed* for ML loads, or if there's another way to support both I missed.
- Još 3 druga odgovora
Novi razgovor -
-
-
@ApacheMADlib was a pioneer in open source in-database machine learning (https://madlib.apache.org/ ). - Još 2 druga odgovora
Novi razgovor -
-
-
If this integrates with
@ApacheSpark SQL then it will make scaling super awesome. -
Make sure to check our {sparklyr} ! https://spark.rstudio.com/mlib/
Kraj razgovora
Novi razgovor -
-
-
I’m finally going through the Andrew Ng course and had this thought along the way. Delighted to find out it already exists!
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
-
-
-
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
-
Čini se da učitavanje traje već neko vrijeme.
Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.