# For those interested in dimensionality reduction

**URL:** <https://discourse.flucoma.org/t/for-those-interested-in-dimensionality-reduction/434>\
**Category:** Learning Resources\
**Created:** [February 19, 2020, 10:17am UTC](https://discourse.flucoma.org/t/for-those-interested-in-dimensionality-reduction/434 "2020-02-19T10:17:17Z")\
**Posts on this page:** 4\
**Page:** 1

<div class="post-metadata">

**Author:** ![tremblap](https://discourse.flucoma.org/user_avatar/discourse.flucoma.org/tremblap/32/5_2.png) [@tremblap](https://discourse.flucoma.org/u/tremblap)\
**Post date:** [February 19, 2020, 10:17am UTC](https://discourse.flucoma.org/t/for-those-interested-in-dimensionality-reduction/434/1 "2020-02-19T10:17:17Z")

</div>

This tutorial is quite impressive!

> **[Idyll](https://idyll-lang.org/gallery/the-beginner-s-guide-to-dimensionality-reduction)**
>
> Idyll Example | The Beginner's Guide to Dimensionality Reduction

---

<div class="post-metadata">

**Author:** ![tremblap](https://discourse.flucoma.org/user_avatar/discourse.flucoma.org/tremblap/32/5_2.png) [@tremblap](https://discourse.flucoma.org/u/tremblap)\
**Post date:** [February 19, 2020, 10:39am UTC](https://discourse.flucoma.org/t/for-those-interested-in-dimensionality-reduction/434/2 "2020-02-19T10:39:14Z")

</div>

BTW, if like me you wonder what is an hyperparameter, I’m told that it is the parameter of the algorithm itself, like the learning rate in a neural net.

---

<div class="post-metadata">

**Author:** ![bafonso](https://discourse.flucoma.org/user_avatar/discourse.flucoma.org/bafonso/32/771_2.png) [@bafonso](https://discourse.flucoma.org/u/bafonso)\
**Post date:** [February 19, 2020, 1:26pm UTC](https://discourse.flucoma.org/t/for-those-interested-in-dimensionality-reduction/434/3 "2020-02-19T13:26:52Z")

</div>

You can read more about this distinction [here](https://datascience.stackexchange.com/questions/14187/what-is-the-difference-between-model-hyperparameters-and-model-parameters). It appears that the consensus is that a hyper parameter is a stricter definition of a parameter that is learned from the data as opposed to being defined by the user. In practice in ML circles I mostly see people mention hyper parameters and not worry so much about the distinction.

---

<div class="post-metadata">

**Author:** ![jamesbradbury](https://discourse.flucoma.org/user_avatar/discourse.flucoma.org/jamesbradbury/32/1252_2.png) [@jamesbradbury](https://discourse.flucoma.org/u/jamesbradbury)\
**Post date:** [February 19, 2020, 1:28pm UTC](https://discourse.flucoma.org/t/for-those-interested-in-dimensionality-reduction/434/4 "2020-02-19T13:28:45Z")

</div>

Awesome! I’m going to send this to my mum. She looked a bit sick last time I tried to explain to her what was going on.