5.7, due Nov 2

Difficult:
I thought the hardest part of this section was the part on marginal distributions. I was confused by the notation in p.m.f. for marginal distributions -- are we summing/integrating over just the variable that we are taking the marginal distribution of? Or are we summing over all of the variables except the variable that we are taking the marginal distribution of?

Reflective:
These are some awesome concepts, and I think it would help a ton to solidify understanding if we had a machine learning algorithm to code up that applied these principles. In Dr. Wingate's CS 401R class we did a lab on the Expectation Maximization algorithm for clustering -- one of the most fundamental machine learning algorithms. That lab applied the concepts of covariance matrices, multivariate normal distributions, and also was relevant in showing the numerical instability of covariance matrices. It was a great lab, and I think ACME should cannibalize it. If you're interested, here's the link: http://liftothers.org/dokuwiki/doku.php?id=cs401r_w2016:lab13

Comments

Post a Comment

Popular posts from this blog

8.7, due December 6

Difficult: The hard part of this section for me was understanding formula 8.48 and the Vector Space Structure section -- specifically the part about scaling relations. Reflective: I can see how these kinds of functions could be better for approximating functions with local behavior, whereas to express such functions in a Fourier basis would be very computationally expensive. I can also see how the simplicity of the function could cause it to be easier to work with (differentiate, evaluate, etc.)
Read more

8.5, due Dec 3

Difficult: I understood the mathematical concept of a convolution. And I understand how convolutional neural networks work like the back of my hand. But after reading this section, I still don't understand how the two mesh together. I was on board with the description given in the intro to 8.5, until they barely mentioned filters and feature extraction again for the rest of the section. Ex. 8.5.11 helped, but more examples like that in class would make a big difference. Reflective: I enjoyed what I did understand of Ex. 8.5.11, which explained that you could apply a convolution to a set of samples to remove some frequencies while leaving others untouched. I still missed the "How", though.
Read more

8.6, due Dec 5

Difficult: This section was actually pretty reasonable. The only part I had trouble with was understanding their method of antialiasing. I don't understand exactly how you can remove all Fourier coefficients outside a certain interval. Reflective: I really liked their examples of aliasing -- the spinning rotor blades, and even the plotting of functions. That was creative but very applicable example.
Read more