# Swarming on anomaly detection mission

**URL:** <https://discourse.numenta.org/t/swarming-on-anomaly-detection-mission/6417>\
**Category:** NuPIC\
**Created:** [August 4, 2019, 8:48pm UTC](https://discourse.numenta.org/t/swarming-on-anomaly-detection-mission/6417 "2019-08-04T20:48:09Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![Moti](https://avatars.discourse-cdn.com/v4/letter/m/87869e/32.png) [@Moti](https://discourse.numenta.org/u/Moti)\
**Post date:** [August 4, 2019, 8:48pm UTC](https://discourse.numenta.org/t/swarming-on-anomaly-detection-mission/6417/1 "2019-08-04T20:48:09Z")

</div>

Hello,  
I build an HTM model that need to find anomaly on my datasets, I am using only with the raw anomaly score.  
My question is, I need to apply swarming mechanism on my model ? or swarming improve and efficient only in classification problem?.  
What the difference in the swarming mechanism between classification and anomaly prediction?

I want to swarm only the HTM model parameters like Cells Per Column, the encoders params I want to define by myself, so using the getScalarMetricWithTimeOfDayAnomalyParams not suitable to my case… in addition, I am working only with the raw anomaly score (prediction cells vs actual cells).

Thanks,  
MAK

---

<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:** [August 5, 2019, 1:38pm UTC](https://discourse.numenta.org/t/swarming-on-anomaly-detection-mission/6417/2 "2019-08-05T13:38:49Z")

</div>

> [@Don't swarm for Anomaly models](https://discourse.numenta.org/t/dont-swarm-for-anomaly-models/915):
>
> Continuing the discussion from [Bad Anomaly detection for complex periods data](https://discourse.numenta.org/t/bad-anomaly-detection-for-complex-periods-data/914): This keeps coming up in different ways, so I just want to clarify. If you want to get temporal anomalies from streams of scalar data, you do not need to run a swarm. We already have model parameters you can use the are generally good at detecting scalar anomalies. You can get them by calling [this function](http://nupic.docs.numenta.org/1.0.3/api/opf/utils.html?highlight=getscalarmetricwithtimeofdayanomalyparams) with the right parameters: Here is an example case of them being used for NAB (thanks @alavin): There you g…

---

<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:** [August 5, 2019, 1:39pm UTC](https://discourse.numenta.org/t/swarming-on-anomaly-detection-mission/6417/3 "2019-08-05T13:39:00Z")

</div>