Proposal Peer Review

[JamesonOates]

---

name: Proposal peer review about: Peer review of final project proposal title: Proposal peer review labels: '' assignees: ''

The following is the peer review of the project proposal by neural_networks. The team members who participated in this review are:

• James Aas - @JamesAas

• Audrey Mills - @audreymmm

• Jameson Oates - @JamesonOates

• Describe the goal of the project.

The project uses neural networks through R studio, and plans to load packages that will classify images based on the brightness of the pixels used in the image.

• Describe the data used or collected.

They are using a dataset that contains 60,000 observations, each representing an image of clothing. Variables examined for each observation include a label for what type of clothing it is, and a column for each pixel in the image, with an associated value for brightness.

• Describe how the research question will be answered, e.g. what approaches/methods will be used.

Neural networks will first be trained to distinguish types of clothing using a dataset consisting of photos of clothing with each pixel characterized for brightness, the project will then be presented using a website and a shiny app to demonstrate to the class how neural networks work.

• Is there anything that is unclear from the proposal?

The proposal was a little bit vague on how one goes about training a neural network. What kind of coding does this entail? We’re also very curious as to how pixel brightness is going to be an indicator of what type of clothing something is, given that clothes are generally a bunch of different shades. Is it based on relative brightness?

• Provide constructive feedback on how the team might be able to improve their project.

It might be useful to find examples of similar neural network research (classifying by brightness) to show the practical use of such networks.

• What aspect of this project are you most interested in and would like to see highlighted in the presentation?

We would love to learn more about how training a neural network works and maybe some information about what specifically they can be used for, real-world applications of this type of data science.

• Provide constructive feedback on any issues with file and/or code organization.

Your proposal is named cartoon character classification, and the repo project organization section of your proposal says that you are working with cartoon characters, but your proposal mainly discusses classifying clothing items. The data itself seems to be for the clothing, so just make sure that you update your project name and anywhere else it might be relevant :)

• (Optional) Any further comments or feedback?

This in an interesting topic of research, very excited to learn more about it!

[elliottchang]

Thanks so much for your feedback! It was very helpful!