<link1> : Couple of sentence summary of paper 1.
<link2> : Couple of sentence summary of paper 2.
. . .
foo1.py : Summary of what you did to foo1.py (or if it is new, what it does) with link to github.
foo2.py : Summary of what you did to foo2.py (or if it is new, what it does) with link to github.
. . .
Section 1 of paper BLAH: if you are writing a paper or report talk about what sections you are workign on and what the status is
Section 2 of paper BLAH: if you are writing a paper or report talk about what sections you are workign on and what the status is
. . .
Figure 1 : Figure 1 with a caption that describes what is being plotted. Add link to code on github that created the plot.
Figure 2 : Figure 2 with a caption that describes what is being plotted. Add link to code on github that created the plot.
. . .
This section should be a discussion of what you did, how you did it, why you did it, and what you found. For example,
In order to explore the extent to which dimensionality reduction plus a classifier can be used to tag each spectrum with a known spectral type, I wrote pca_classify.py (link to code) which first whitened the spectra and then used scikit-learn’s PCA functionality to decompose the sample into N=20 principle components. The first 5 are shown in Figure 1. The figure shows that these components are dominated by only a few spectral lines. Figure 2 shows the explained variance as a function of the number of included principle components and suggests that N=20 is insufficient. Nevertheless, by projecting onto these components and training an SVM classifier (using grid search on the parameters Cand ; see Figure 3) with training and testing sets of size 3,500 and 1,500 respectively, I achieved an accuracy of 80% on the testing set.
List the tasks that you will perform next week. For example: perform PCA on power spectra sample, investigate the PCA components and identify the one(s) that could be associated to instrumental signal.
In this section it would be inapprioriate, following the example above, to continue describing the tasks as in: then perform reconstruction of power spectra without that component and recreate the original time series witouth the instrumental signal unless you have some resonable confidence that you will be able to perform all 4 tasks. In other words, from week to week you are responsible for performing the tasks you propose to perform, or report on roadblocks that prevented their execution.