Applying dimensional reduction techniques to Kepler data.
Here we give an example of the dimensional reduction techniques using jupyter notebooks. We have 3 different methods we try in this notebook:
With these three we hope to visualize distributions of avalanches of stars. In particular, we want to visualize the space of such distributions. Since this space is approximately 500-dimensional, we use the three different dimensional reduction techniques to visualize our results.
In order to understand the background behind this project, please see this blog post:
https://publish.illinois.edu/mohammedsheikh/2017/09/23/hello-world-2/
In order to look at the LEM algorithm, please take a look at this blog post:
https://publish.illinois.edu/mohammedsheikh/2017/11/30/manifold-learning/
To see the mathematical intuition behind manifold learning, see this blog post:
https://publish.illinois.edu/mohammedsheikh/2017/11/30/the-geometry-behind-manifold-learning/
Finally, to see comparison of TSNE and LEM, take a look at this blog post:
https://publish.illinois.edu/mohammedsheikh/2017/12/04/evaluation-of-lem-and-t-sne/