Life Complete

Week 13: 4/18-4/22

What did I do this week?

This week, I wrapped up Life. The project itself wasn’t too hard to figure out conceptually, but we ran into issues as to how to write a handler class. Conceptually, the handler class made sense, but we had to get through the only enough memory for the super class hurdle. Eventually, we got some pretty good hints from Downing’s schedule where he released working code for an all-purpose handle class. That was really helpful.

What’s in my way?

Nothing much. Only the final OOP test is looming.

What will I do next week?

I will begin studying for the final test. It’s a big portion of my grade, so I don’t want to mess that up.

Thoughts

This week we really discussed OO design. I really enjoyed the lectures as they were conceptually stimulating. A lot of the discussion was more of “why” instead of “how.” Explaining why a certain design is preferable to another is something that has to make sense and must be convincing for the other person. The reasons talked about in class were convincing and useful for convincing not only others but also myself that the design was sound.

Recent Discovery

In light of the recent breakthroughs of Convolutional Neural Networks, this paper highlights some disadvantages of these CNNs and how they can be TOO good at locating features that aren’t actually there.

Written on April 18, 2016