Space Science with Python - AI 1-12: Autoencoder Latent Space Visualization

Viewers! In this second Autoencoder video we will take a look at the latent space of our asteroid spectra. Based on our architecture, we compressed the 49-dimensional spectra to a 2-dimensional latent domain and will take a look at the resulting data.

We will create an interactive visualizer using ipywidgets to display the 2-D data points and overlay it with the corresponding spectra classes.

Will the asteroid classes be distinguishable in low-dim? Let's see!

---

---

Content
0:00 Introduction / Recap
1:25 Cell #1 to #9 - Recap from last time
2:20 Cell #10 - Pandas dataframe with all results
3:55 Cell #11 - Latent space visualization (Matplotlib)
7:33 Cell #12 - Interactive latent space visualization (ipywidgets)
12:00 Summary & Outlook

---

There is a lot to do and to learn and I hope you will join the journey. Meanwhile, if you have questions or ideas, reach out to me via:

Or drop a comment!

Talk to you later,
Thomas