# Train an integer instead of a classification

**URL:** <https://community.konduit.ai/t/train-an-integer-instead-of-a-classification/616>\
**Category:** Uncategorized\
**Created:** [June 15, 2020, 6:34am UTC](https://community.konduit.ai/t/train-an-integer-instead-of-a-classification/616 "2020-06-15T06:34:29Z")\
**Posts on this page:** 1\
**Showing post:** 2

<div class="post-metadata">

**Author:** ![treo](https://yyz1.discourse-cdn.com/flex035/user_avatar/community.konduit.ai/treo/32/47_2.png) [@treo](https://community.konduit.ai/u/treo)\
**Post date:** [June 15, 2020, 6:49am UTC](https://community.konduit.ai/t/train-an-integer-instead-of-a-classification/616/2 "2020-06-15T06:49:30Z")

</div>

You can’t train an integer, but you can train a float.

As I already said in a [previous answer](https://community.konduit.ai/t/regression-instead-of-a-classification/595), you change the classification part of the record reader to be regression.

And then you change the output layer to use LossMSE with Identity activation.

And obviously you will want to apply [regression evaluation](https://javadoc.io/static/org.deeplearning4j/deeplearning4j-nn/1.0.0-beta7/org/deeplearning4j/nn/multilayer/MultiLayerNetwork.html#evaluateRegression-org.nd4j.linalg.dataset.api.iterator.DataSetIterator-) instead of the normal classification regression.

---

_[View the full topic](https://community.konduit.ai/t/train-an-integer-instead-of-a-classification/616)._