# Mathematica and ML – TensorFlow visualisation

**URL:** <https://community.wolfram.com/t/mathematica-and-ml-tensorflow-visualisation/21551>\
**Category:** Questions\
**Tags:** data-science, graphics-and-visualization, machine-learning\
**Created:** [February 4, 2023, 4:58pm UTC](https://community.wolfram.com/t/mathematica-and-ml-tensorflow-visualisation/21551 "2023-02-04T16:58:06Z")\
**Posts on this page:** 4\
**Page:** 1

<div class="post-metadata">

**Author:** ![marknorgate](https://avatars.discourse-cdn.com/v4/letter/m/ad7895/32.png) [@marknorgate](https://community.wolfram.com/u/marknorgate)\
**Post date:** [February 4, 2023, 4:58pm UTC](https://community.wolfram.com/t/mathematica-and-ml-tensorflow-visualisation/21551/1 "2023-02-04T16:58:06Z")

</div>

I am currently studying a specialization on Coursera in Machine Learning and am investigating various tools to help me out with the maths and with visualisations and so on.

Although I have many decades experience of development and am very handsome, I haven’t done any formal maths study since my early 20s. So I want to take the edge off.

As a first excursion, I would like to be able to plot the results of predictions on my TensorFlow models against the training data, in a similar way `tfvis` library that comes along with `tensorflow.js`; but perhaps allowing me to tweak the layers, units and so on of my regressor model and see the effect in real time.

Am I dreaming here? Can Mathematica help me with that? If not, can anyone suggest an alternative?

---

<div class="post-metadata">

**Author:** ![wolframcommunity](https://yyz1.discourse-cdn.com/wolfram/user_avatar/community.wolfram.com/wolframcommunity/32/8335_2.png) [@wolframcommunity](https://community.wolfram.com/u/wolframcommunity)\
**Post date:** [February 5, 2023, 4:02pm UTC](https://community.wolfram.com/t/mathematica-and-ml-tensorflow-visualisation/21551/2 "2023-02-05T16:02:03Z")

</div>

Welcome to Wolfram Community! Please provide your trials in terms of Wolfram Language code. This will make it easier for other members to help you.

Check several methods available to include your code in the rules [http://wolfr.am/READ-1ST](http://wolfr.am/READ-1ST)

---

<div class="post-metadata">

**Author:** ![marknorgate](https://avatars.discourse-cdn.com/v4/letter/m/ad7895/32.png) [@marknorgate](https://community.wolfram.com/u/marknorgate)\
**Post date:** [February 5, 2023, 5:03pm UTC](https://community.wolfram.com/t/mathematica-and-ml-tensorflow-visualisation/21551/3 "2023-02-05T17:03:24Z")

</div>

Thank you for your response, but to be honest I haven’t even made the attempt yet.

I am asking broadly if this is a thing people use the platform for? If not, I don’t want to spend time investigating. If so, I shall apply effort.

That might sound lazy, but I can short-circuit if it doesn’t do that. Just a very high-level “does it do this”?

---

<div class="post-metadata">

**Author:** ![jasonb](https://yyz1.discourse-cdn.com/wolfram/user_avatar/community.wolfram.com/jasonb/32/20741_2.png) [@jasonb](https://community.wolfram.com/u/jasonb)\
**Post date:** [February 6, 2023, 2:38pm UTC](https://community.wolfram.com/t/mathematica-and-ml-tensorflow-visualisation/21551/4 "2023-02-06T14:38:03Z")

</div>

> That might sound lazy

it does.

> I have many decades experience of development and am very handsome

Well that might help - you know what they say, if you are going to be lazy at least do it handsomely.