Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Paper: Deep and Ensemble Learning to Win the Army RCO AI Signal Classification Challenge #473

Open
wants to merge 9 commits into
base: 2019
from

Conversation

Projects
None yet
5 participants
@Teque5
Copy link

commented May 22, 2019

This pull request is for 2of2 papers from The Aerospace Corporation approved for submission to the 2019 SciPy Conference.

OTR201900665 for the paper Deep and Ensemble Learning to Win the Army RCO
AI Signal Classification Challenge

Everything built OK on my box.

Deep and Ensemble Learning Paper
The following squashed commit was approved for public release by **The
Aerospace Corporation** on 2019-05-20. Commits made by the *Digital
Communications Implementation Department* are approved for public
release under request #OTR201900733 by the Office of Technical relations.
@jbednar
Copy link

left a comment

Looks good. A nice overview of the approach, tools used, and results obtained.

Show resolved Hide resolved papers/aero_nn_vila/aero_nn_vila.rst Outdated
Show resolved Hide resolved papers/aero_nn_vila/aero_nn_vila.rst Outdated
Show resolved Hide resolved papers/aero_nn_vila/aero_nn_vila.rst Outdated
Show resolved Hide resolved papers/aero_nn_vila/aero_nn_vila.rst Outdated
Show resolved Hide resolved papers/aero_nn_vila/aero_nn_vila.rst Outdated
Show resolved Hide resolved papers/aero_nn_vila/aero_nn_vila.rst Outdated
Show resolved Hide resolved papers/aero_nn_vila/aero_nn_vila.rst Outdated
Show resolved Hide resolved papers/aero_nn_vila/aero_nn_vila.rst Outdated
Show resolved Hide resolved papers/aero_nn_vila/aero_nn_vila.rst Outdated
Show resolved Hide resolved papers/aero_nn_vila/aero_nn_vila.rst Outdated
@deniederhut

This comment has been minimized.

Copy link
Member

commented May 27, 2019

Hm... it looks like you've marked James' suggestions as "resolved" but I don't see that any of them have been addressed, either in the comments or in changes to the paper. Maybe there is a commit that didn't get pushed to GitHub?

@Teque5

This comment has been minimized.

Copy link
Author

commented May 29, 2019

@deniederhut I have updated my pull request and merged his suggestions.

@shuaitang

This comment has been minimized.

Copy link

commented Jun 10, 2019

Typo: datset - dataset

Concerns:
1/ "While the best architectures turned out to be ResNet and ResNeXt, the authors don’t believe there is anything inherent in those architectures that makes them more suited to the modulation-classification problem."

It is not a scientific way of arguing, and it is not supported by any observable or experimented evidence. Please revise this sentence in your conclusion.

2/ "This paper also showed an innovative method of merging different neural networks that were trained with significantly different data. "

I am not sure if it is novel as feature-fusion and multi-view learning exist for a while now in machine learning and deep learning. Authors should check these two research areas and see if necessary references should be included.

@jbednar

This comment has been minimized.

Copy link

commented Jun 10, 2019

It is not a scientific way of arguing, and it is not supported by any observable or experimented evidence. Please revise this sentence in your conclusion.

The wording could perhaps be improved; it's true that such a claim is unsubstantiated, with no evidence in the paper to justify it. Still, I do appreciate having the authors' best judgment and candor to help us interpret the results. It's worse to leave the false impression that the authors are claiming that those particular architectures are somehow crucial or vitally important, and so I'd favor finding some way to continue to express this opinion in the text.

@andresvila

This comment has been minimized.

Copy link

commented Jun 10, 2019

Our latest PR doesn't address the comments yet.

We will rephrase the first comment and leave the information in the paper. Thanks for all the comments about it.

We will address the second comment by making it more specific to the actual method used to fuse the classifiers. We do not intend claim to be the first to merge classifiers.

@Teque5

This comment has been minimized.

Copy link
Author

commented Jun 12, 2019

I just pushed the latest with the above changes as discussed.

@deniederhut

This comment has been minimized.

Copy link
Member

commented Jun 13, 2019

It looks like at least one of these papers is missing its DOI. Could I ask you to take a stroll through you bibliography and add these?

@Teque5

This comment has been minimized.

Copy link
Author

commented Jun 17, 2019

@deniederhut DOIs have been added/updated. Good catch.

@deniederhut

This comment has been minimized.

Copy link
Member

commented Jun 18, 2019

Thanks for adding those 😄

Hey @jbednar and @shuaitang ! Do you feel this paper is now ready for inclusion?

@jbednar

This comment has been minimized.

Copy link

commented Jun 18, 2019

Sure, I'm happy with it!

@shuaitang

This comment has been minimized.

Copy link

commented Jun 18, 2019

It looks good to me. Thanks!

@deniederhut

This comment has been minimized.

Copy link
Member

commented Jun 20, 2019

Excellent! Thanks for submitting to SciPy!

@deniederhut deniederhut added the ready label Jun 20, 2019

@Teque5

This comment has been minimized.

Copy link
Author

commented Jun 21, 2019

@deniederhut For the poster (and other materials) should I assume there will be PDFs of these papers hosted at http://conference.scipy.org/proceedings/scipy2019/ by the time of the conference?

@deniederhut

This comment has been minimized.

Copy link
Member

commented Jun 22, 2019

Yup! We'll have the conference-ready proceedings up online before the conference. Afterward, we put in links to all the talks (they get posted on YouTube) and we re-publish with the video links and DOIs.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
You can’t perform that action at this time.