-
Notifications
You must be signed in to change notification settings - Fork 6
/
Copy pathanswers.txt
64 lines (43 loc) · 5.28 KB
/
answers.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
Tell us about the features and technologies you used in your Swift playground.
--
The goal of my Swift playground was to create a neural network class that supported basic neural network operations. I wanted to show the power of neural networks and how they work without getting too technical. Even though iOS has some built in neural network libraries, I stayed away from them to prove it wasn’t a black box or super complex. I also thought it would be cool if I just did it myself.
My playground uses Foundation, UIKit, CoreGraphics, and PlaygroundSupport. On top of those frameworks and standard Swift operations, I built a matrix class. Using that matrix class, I built a neural network class. I also used the MNIST dataset (60,000 labeled images of handwritten digits) to train a neural network to detect handwritten digits. Collecting this data myself would’ve been outside the scope of the project.
Additionally, I used the markup formatting that Swift playgrounds can render which made it easier to explain what was going on in the Live View. Because my project focuses on a very specific technology, I needed to take advantage of this to ensure that the user knew what was going on. I tried making it as relevant as possible, but this was difficult because I wasn’t sure how technical my audience would be.
Finally, using UIKit, CoreGraphics, and my neural network class, I created an experience by adding interactivity and showing some real world examples. This was just as important as all the implementation details. This consisted mainly of building graphs and canvases to draw on.
To answer the question, my project consisted mainly of:
- Foundation / Standard programming operations
- OperationQueue / DispatchQueue
- FileHandle / Bundle / Data (for loading weights and images)
- UIKit
- UIImage / UIImageView (for MNIST images)
- UILabel
- UIViewController
- UIView (including touch events)
- UITapGestureRecognizer
- Auto Layout
- UIColor
- UIProgressView
- UINavigationController
- UIBezierPath (for drawing on views)
- UIButton
- CoreGraphics
- CGFloat / CGRect
- PlaygroundSupport
- PlaygroundPage
...and probably a couple more trivial ones.
In what ways have you considered sharing your coding knowledge and enthusiasm for computer science with others?
--
I regularly encourage others to try programming. One way I’ve done this in my local community was by helping build the high school computer science curriculum. I’ve also encouraged my friends to try programming, advising them to work on whatever they think is interesting. When doing this, I try to make it clear that it doesn’t have to be torture to learn how to code! I believe the best way to learn is to simply work on projects that interest you. This will make learning a fun experience rather than a burden.
I also try to stay up to date with the latest technologies and encourage others to join me. I recently did this by offering a few friends to take an online class with me (“Machine Learning” on Coursera). We each watched the videos online on our own time, and met and discussed twice a week. This worked out very well and gave us all some exposure to machine learning.
Another way I share my coding knowledge and enthusiasm is by contributing to the open source community. I have lots of open source repositories on Github and always hope they are helping people learn.
I’m also considering making this playground into a short book. When I was first learning about neural networks, I was actually looking for something similar to this. I wanted a resource to explain the details of neural networks, and then allow me to interactive with them. This is especially applicable with neural networks because there are so many hyper-parameters that affect the way they learn and perform. I think taking this playground and focusing it on teaching people how neural networks work could benefit a lot of people.
If you have extenuating circumstances that would require travel assistance in addition to the lodging already provided for scholarship recipients, let us know.
--
I can handle the traveling costs.
Is there anything else you would like us to know?
--
I have tested this playground on an iPad Pro and a MacBook Pro Retina, but it should work on other devices.
Not knowing my target audience was the most difficult part of this project. I wasn’t sure how technical the person reviewing my playground would be, so I had to take a guess and try and strike the right balance. I hope it isn’t too technical or not technical enough! One way I handled this was by making the last page “optional”, because it is more technical than the other pages.
Also, the neural network and matrix implementations consist mainly of math. I’m not sure if this is considered a “feature or technology,” so I decided to leave it out of the first question even though it’s a large part of my project. The neural network implementation was by far the most math heavy, and consisted of lots of calculus (training a neural network is pretty much just the derivative chain-rule repeated many times).
I have nearly every Apple product. I’ve been doing iOS programming for about 4 years now. I’ve been watching videos from the WWDC since 10th grade. I love Apple, and would really appreciate if I could go to the WWDC this year!
Thanks for this opportunity!