# Deleted data used in model?

**URL:** <https://discuss.elastic.co/t/deleted-data-used-in-model/186488>\
**Category:** Elasticsearch\
**Tags:** elastic-stack-machine-learning\
**Created:** [June 19, 2019, 2:17pm UTC](https://discuss.elastic.co/t/deleted-data-used-in-model/186488 "2019-06-19T14:17:45Z")\
**Posts on this page:** 5\
**Page:** 1

<div class="post-metadata">

**Author:** ![Bharath\_Kumar\_R](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/bharath_kumar_r/32/45474_2.png) [@Bharath\_Kumar\_R](https://discuss.elastic.co/u/Bharath_Kumar_R)\
**Post date:** [June 19, 2019, 2:17pm UTC](https://discuss.elastic.co/t/deleted-data-used-in-model/186488/1 "2019-06-19T14:17:45Z")

</div>

Consider I have created a job today. We will be storing data for only 60 days. On the 61st day, when data is fed to the job, will the job create a new model with the newly fed data or will the job store the past historical data related to the model(and continue modelling newly fed data using the past deleted data)?

---

<div class="post-metadata">

**Author:** ![richcollier](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/richcollier/32/115035_2.png) [@richcollier](https://discuss.elastic.co/u/richcollier)\
**Post date:** [June 19, 2019, 4:22pm UTC](https://discuss.elastic.co/t/deleted-data-used-in-model/186488/2 "2019-06-19T16:22:36Z")

</div>

The model is cumulative and only needs to see the raw data once, hence on the 61st day, only that data from the 61st day is used by the job. Therefore the maturity of the model can be greater than the retention policy of the index the job is operating on.

---

<div class="post-metadata">

**Author:** ![Bharath\_Kumar\_R](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/bharath_kumar_r/32/45474_2.png) [@Bharath\_Kumar\_R](https://discuss.elastic.co/u/Bharath_Kumar_R)\
**Post date:** [June 20, 2019, 10:11am UTC](https://discuss.elastic.co/t/deleted-data-used-in-model/186488/3 "2019-06-20T10:11:13Z")

</div>

So, what you are saying is the job won't create a model for new data based upon past deleted data but only based on current data, right?

---

<div class="post-metadata">

**Author:** ![richcollier](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/richcollier/32/115035_2.png) [@richcollier](https://discuss.elastic.co/u/richcollier)\
**Post date:** [June 20, 2019, 10:26am UTC](https://discuss.elastic.co/t/deleted-data-used-in-model/186488/4 "2019-06-20T10:26:33Z")

</div>

What I'm saying is that the model is created from observing all historical data that fed to the job. The model is cumulative and the data only needs to be seen once. When new data comes in the new data updates the model and the model adapts overt ime

---

<div class="post-metadata">

**Author:** ![system](https://us1.discourse-cdn.com/elastic/original/3X/1/a/1ac57faf039f6b580b3f104ef42a2a89e41014de.png) [@system](https://discuss.elastic.co/u/system)\
**Post date:** [July 18, 2019, 10:26am UTC](https://discuss.elastic.co/t/deleted-data-used-in-model/186488/5 "2019-07-18T10:26:39Z")

</div>

Closed on July 18, 2019, 10:26am UTC

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.