rostlab / js16_projectb_group6 Goto Github PK

Is there a name for your algorithm that we can use when presenting it on the website?

@gyachdav @goldbergtatyana @sacdallago
As we already mentioned here #60 (comment)
there are some inconsistencies between the book and the tv show. project a uses data from the awoiaf wiki which uses the book data in the info box (our main source of information). meanwhile, we're doing a prediction for the tv show and some characters that are dead in the tv show are still alive in the books and vice versa. a notable example that also shows up in the top 10 popular plod predictions is Stannis Baratheon who was killed by Brienne in the TV show but is still very much alive in the books. Also, the books contain much more characters which, as a sum, probably influenced our plod predictions as well. Should we just leave it like that then?

Add package.json

@sacdallago Should we change the populartity (page rank) to be taken from the api?
only asking because feature freeze was on monday.

• provide call to rerun SVM
• implement getPlodByCharacter
• implement getPlodOrdered limit by given number
• implement getPlod ordered of all characters
• implement getFeatureContribution

Create includable functions that get data from the API from project a

• getAllCharacters
• getAllAllegiances
• getAllCultures
• getAllTitles

Please provide your NPM package as soon as possible!

From the Online Oxford Dictionary:

Protected against attack or other criminal activity

https://github.com/Rostlab/JS16_ProjectB_Group6/blob/develop/.travis.yml#L23

Did you provide the function returning PLODs to group A? Also, please upload the file with the final predictions for all characters here and in your model description (so that we don't need to scroll up and down through old posts to find the correct file). What are your top 10 characters, please list them here as well. Thanks!

Hello there, people. As you should have noticed, tests are failing 😸
Reason is directory structure has changed, so tests can't "require" modules.
So the question is : Is this the final directory structure of the project?

In an effort to bring together the pieces, we came up with the following requirements:

The predictions that you will produce will be bound to a character, say

{
  name: "Some character",
  ...,
  PLOD6: //YOUR NORMALIZED PLOD HERE,
  PLOD7: // B7's PLOD here
}

As the PLOD for a character will be a numerical value and the machine learning will only train on the available data now, so it makes no sense for you to create PLODs dynamically.

You will write directly on the database that A uses that should be somewhere on mLab. If @kordianbruck is using his own server to store the data, please come up with some idea.

@AlexMoroz I assigned the issue to you because E should supervise this, but defer commitment to one of your teammates.

@Hack3l you commited on the master branch (d3a05d5), this is not what you should do.

Please:

• Revert your commit and bring master back to the initial state ( 72df47b )
• Open a branch for the feature you work on, branching off from develop
• If you are already done with your feature, open a pull request

We would like you to implement the feature model. If you don't know how this works, please have a look at https://www.atlassian.com/git/tutorials/comparing-workflows/feature-branch-workflow/ and the video in which I explain it https://youtu.be/twKtSfLcKxI?t=12m25s

Add jshint

Kernel: 1. polykernel
2. rbfkernel
Attributes: Removed attributes placeOfBirth, placeOfDeath, placeOfLastVisit and age
Parameters: Default

3.attributeranking.txt
3.polykernel.txt
3.rbfkernel.txt.gz

Meysters, as part of the media blitz we're planning there will be a press release that will throw some *_big numbers *_at the readers. Can you provide some impressive statistics about the data you processed to come up with your predictions. something like looked at 25 features for 2000 characters totaling in 500k data points. Any thing that you think might be interesting IS interesting.

• implement getPLOD
• implement getFeatureRanking
• cleanup files

Add Travis configuration

Hey guys,
every endpoint must be configurable. This package needs to be independent of everything else, also of bruck's server!

Look how C10 implemented this: https://github.com/Rostlab/JS16_ProjectC_Group10/wiki/Init-New-GoT-Map-and-Options

• Pull data from database
• Convert data to arff
• Add more features
• Test features in svm
• Create list of feature contribution

• Find good svm package
• Implement SVM
• Find best settings for SVM
• Test SVM with selected features
• Store plod and feature contribution

Compare the SVMs with and without books, to decide if they are a good feature.