Message Boards Message Boards

1
|
2747 Views
|
2 Replies
|
3 Total Likes
View groups...
Share
Share this post:

Classify & Predict: How to extract imputed data?

Posted 1 year ago

The functions "Classify" and "Predict" are working with an 'embedded' preprocessing pipeline for data preparation. The following image from Etienne Barnard's book "Introduction to Machine Learning" shows what is meant.

Screenshot

One step of the preprocessing pipeline is to Synthesize Missing Values which finally means the function imputes missing data by itself.

My question: How to get access / how to extract the imputed data?

POSTED BY: Jürgen Kanz
2 Replies

Using the same dataset

data = ResourceData["Sample Data: Titanic Survival"];
titanic = Classify[data -> "SurvivalStatus"]

after training you can extract the estimated data distribution using Information

dist = Information[titanic, "MissingSynthesizer"]

then it is possible to recreate the imputed data separately

imputed = SynthesizeMissingValues[dist, data]

Giulio,

Thank you. Your solution works well.

Have a nice weekend.

POSTED BY: Jürgen Kanz
Reply to this discussion
Community posts can be styled and formatted using the Markdown syntax.
Reply Preview
Attachments
Remove
or Discard

Group Abstract Group Abstract