# inferenceType: TemporalClassification

**URL:** <https://discourse.numenta.org/t/inferencetype-temporalclassification/2846>\
**Category:** NuPIC\
**Created:** [September 18, 2017, 2:02pm UTC](https://discourse.numenta.org/t/inferencetype-temporalclassification/2846 "2017-09-18T14:02:24Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![dervinism](https://avatars.discourse-cdn.com/v4/letter/d/e19adc/32.png) [@dervinism](https://discourse.numenta.org/u/dervinism)\
**Post date:** [September 18, 2017, 2:02pm UTC](https://discourse.numenta.org/t/inferencetype-temporalclassification/2846/1 "2017-09-18T14:02:25Z")

</div>

Is there an HTM implementation for classifying temporal sequences? I came across inferenceType: TemporalClassification listed in the parameters file for creating an HTM network example. Is this option functioning or is it only provisional yet?

---

<div class="post-metadata">

**Author:** ![rhyolight](https://yyz2.discourse-cdn.com/flex030/user_avatar/discourse.numenta.org/rhyolight/32/3922_2.png) [@rhyolight](https://discourse.numenta.org/u/rhyolight)\
**Post date:** [September 18, 2017, 2:56pm UTC](https://discourse.numenta.org/t/inferencetype-temporalclassification/2846/2 "2017-09-18T14:56:16Z")

</div>

@marion did a bunch of work on temporal classification, but I don’t think she had much success. It is a hard problem (it is basically temporal pooling). See [How SMI might inform Temporal Pooling](https://discourse.numenta.org/t/how-smi-might-inform-temporal-pooling/2515).

---

<div class="post-metadata">

**Author:** ![jacobeverist](https://yyz2.discourse-cdn.com/flex030/user_avatar/discourse.numenta.org/jacobeverist/32/6736_2.png) [@jacobeverist](https://discourse.numenta.org/u/jacobeverist)\
**Post date:** [September 18, 2017, 6:56pm UTC](https://discourse.numenta.org/t/inferencetype-temporalclassification/2846/3 "2017-09-18T18:56:43Z")

</div>

I believe @marion’s work is located here:

> **[numenta/htmresearch](https://github.com/numenta/htmresearch/tree/master/projects/sequence_classification)**
>
> htmresearch - Experimental algorithms. Unsupported.

There’s a larger body of research out there on sequence classification in general with lots of different techniques such as Dynamic Time Warping and various types of distance metrics.