noblekennamer / astro_porject Goto Github PK

@abhisaarsharma @arielkruger1989 @dcockerham

Hey everyone the report is up. I wrote the intro, results and analysis, and conclusion.

At the end of the report there are 3 sections:

1. What we did
2. What we used (libaries, papers, etc...)
3. our personal experience about the project

If everyone could fill those sections directly underneath their name that would be great. If we could get that part done by tomorrow that would be best so that we can finalize the rest of the report after that.

Also I've never worked with word docs on git before so just to be safe it might worthwhile to make a copy of the document and put your changes in red. I can then go and paste them in.

Just make an issue whenever you change something and I'll go change it.

Write now the report is almost 20 pages so we are meeting the minimum page count of 7 pretty easily.

@abhisaarsharma

Build an ANN to classify stars. @NobleKennamer

Hey @arielkruger1989 @abhisaarsharma @dcockerham check out the link below

http://scikit-learn.org/stable/auto_examples/classification/plot_classifier_comparison.html

I added in example for splitting the dataset, training a model, and plotting an roc curve.

@dcockerham @arielkruger1989 @abhisaarsharma

@dcockerham

@arielkruger1989

@dcockerham

How is the poster coming? Do you need any help?

Build a decision tree to classify stars. @arielkruger1989

@dcockerham

@abhisaarsharma

@arielkruger1989 @abhisaarsharma @dcockerham

When can people meet this weekend. preferably Saturday.

Just post anytime on here.

I'm free all day

@arielkruger1989 @abhisaarsharma @dcockerham

If people could document the main notebooks RR_Lyrae.ipynb and redshift.ipynb so that the graders can easily follow everything that is happening that would be great.

Instead of adding comments into the executable cells create markdown cells directly above them to add the documentation. That looks a lot nicer. Plus the markdown cells support HTML so we can really make them look nice.

a good tutorial for adding this markdown can be found in the link below(it is actually really easy)

http://jupyter.cs.brynmawr.edu/hub/dblank/public/Jupyter%20Notebook%20Users%20Manual.ipynb

Build gaussian process to classify stars @NobleKennamer

@dcockerham

I've been playing around with these methods and trying to get them to produce usable results, but with zero success. As best I can figure, they simply aren't suited to a classification problem like this. After fitting, they all seem to end up predicting a zero for any input (this does, amusingly, give them good accuracy due to the sparseness of RR Lyrae stars, but is obviously a useless result).

@NobleKennamer. Noble I made a copy of the notebook just to make sure I would not mess with the previous one.
In this notebook I optimized random forest. Added new model for boosted decision trees and fixed my implementation of decision trees. If it is everything correct then I can show the changes I made in the files or we can just replace the old file. Let me know what you think is better for us to do.

@dcockerham , @arielkruger1989 , @abhisaarsharma

Send me any sources you use to do your work. I'm maintaining the bibliography as we go.

@arielkruger1989 @abhisaarsharma @dcockerham

The project has been finished and turned.

Please do not update this GitHub so that it matches exactly what I turned in. I pointed the graders to this GitHub so it is important that it stays exactly as is.

@dcockerham Check this out

http://scikit-learn.org/stable/modules/cross_validation.html