madlib-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Robert Glithero <rglith...@pivotal.io>
Subject Re: Apache MADlib Community Call on Thursday, May 10
Date Tue, 08 May 2018 17:30:11 GMT
 Reminder -  please join us for the Apache MADlib Community Call on
Thursday, May 10, at 1100 PDT / 1800 GMT.  We'll discuss and demonstrate
some of the new features in release 1.14

Webinar link:  https://pivotal.zoom.us/j/994236734

New features:
* Balanced datasets - sampling algorithm to balance classification datasets
by resampling using various techniques including undersampling,
oversampling, uniform sampling or user-defined proportion sampling
* Personalized PageRank - new variant of PageRank graph algorithm which is
better for personalization
* Mini-batch optimizer for multilayer perceptron neural networks (and
associated pre-processor function) - faster convergence, more accurate
* PostgreSQL 10.2 support

Improvements:
* K-nearest neighbors - Added weighted averaging/voting by distance.
* Summary - Added more statistics including number of positive, negative,
zero values and 95% confidence intervals.
* Multilayer perceptron - Added support for one-hot encoded categorical
dependent variable for classification.
* Lotsa bug fixes

Release notes:
https://github.com/apache/madlib/blob/master/RELEASE_NOTES

Available now on PivNet.

Thanks to the advanced analytics team for another great release!



Bob Glithero | Data Product Marketing
Pivotal, Inc.
rglithero@pivotal.io | m: 415.483.5220



On Fri, May 4, 2018 at 2:33 PM, Robert Glithero <rglithero@pivotal.io>
wrote:

> Please join us for the Apache MADlib Community Call on Thursday, May 10,
> at 1100 PDT / 1800 GMT.  We'll discuss and demonstrate some of the new
> features in release 1.14
>
> Webinar link:  https://pivotal.zoom.us/j/994236734
>
> New features:
> * Balanced datasets - sampling algorithm to balance classification
> datasets by resampling using various techniques including undersampling,
> oversampling, uniform sampling or user-defined proportion sampling
> * Personalized PageRank - new variant of PageRank graph algorithm which is
> better for personalization
> * Mini-batch optimizer for multilayer perceptron neural networks (and
> associated pre-processor function) - faster convergence, more accurate
> * PostgreSQL 10.2 support
>
> Improvements:
> * K-nearest neighbors - Added weighted averaging/voting by distance.
> * Summary - Added more statistics including number of positive, negative,
> zero values and 95% confidence intervals.
> * Multilayer perceptron - Added support for one-hot encoded categorical
> dependent variable for classification.
> * Lotsa bug fixes
>
> Release notes:
> https://github.com/apache/madlib/blob/master/RELEASE_NOTES
>
> Available now on PivNet.
>
> Thanks to the advanced analytics team for another great release!
>
>
> Bob Glithero | Data Product Marketing
> Pivotal, Inc.
> rglithero@pivotal.io | m: 415.483.5220
>
>
>

Mime
View raw message