# How intent classification works in tensorflow pipeline?

**URL:** <https://forum.rasa.com/t/how-intent-classification-works-in-tensorflow-pipeline/3551>\
**Category:** Rasa Open Source\
**Created:** [December 14, 2018, 6:17am UTC](https://forum.rasa.com/t/how-intent-classification-works-in-tensorflow-pipeline/3551 "2018-12-14T06:17:10Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![goelishankgoel](https://dub1.discourse-cdn.com/flex013/user_avatar/forum.rasa.com/goelishankgoel/32/1008_2.png) [@goelishankgoel](https://forum.rasa.com/u/goelishankgoel)\
**Post date:** [December 14, 2018, 6:17am UTC](https://forum.rasa.com/t/how-intent-classification-works-in-tensorflow-pipeline/3551/1 "2018-12-14T06:17:10Z")

</div>

Tensorflow embeddings are created from the training text but how does it classify multiple intents ? which classifier does it train ? SVM or a it just trains a Neural network or LSTM directly for classification ?

---

<div class="post-metadata">

**Author:** ![JiteshGaikwad](https://avatars.discourse-cdn.com/v4/letter/j/77aa72/32.png) [@JiteshGaikwad](https://forum.rasa.com/u/JiteshGaikwad)\
**Post date:** [December 14, 2018, 6:40am UTC](https://forum.rasa.com/t/how-intent-classification-works-in-tensorflow-pipeline/3551/2 "2018-12-14T06:40:00Z")

</div>

hey @goelishankgoel, as far as what i can understand from the code it uses EmbeddingIntentClassifier,

> <https://github.com/RasaHQ/rasa_nlu/blob/master/rasa_nlu/classifiers/embedding_intent_classifier.py>