{ "id": 2300, "title": "Submission Instructions??", "url": "/competitions/Raising-Money-to-Fund-an-Organizational-Mission/discussion/2300", "author_name": "", "post_date": "2012-08-02T20:22:46.120Z", "votes": null, "comment_count": 3, "views": 1876, "content": "

Hi,

\r\n

On the "Make submission" page it says:

\r\n

Your entry must:

\r\n\r\n

Are the first 2 columns GROUP, ID followed by PREDICTED_AMOUNT2?

\r\n

Does the submission need to be sorted in any particular order?

\r\n

 

\r\n

Also, how can I create variables GROUP & ID for the training set. I ask this since I have validation set on which I would like to test my model before making a leaderboard submission.

", "messages": [ { "id": "12888", "postDate": "08/02/2012 20:22:46", "content": "

Hi,

\r\n

On the "Make submission" page it says:

\r\n

Your entry must:

\r\n\r\n

Are the first 2 columns GROUP, ID followed by PREDICTED_AMOUNT2?

\r\n

Does the submission need to be sorted in any particular order?

\r\n

 

\r\n

Also, how can I create variables GROUP & ID for the training set. I ask this since I have validation set on which I would like to test my model before making a leaderboard submission.

", "rawMarkdown": "", "votes": null }, { "id": "12890", "postDate": "08/02/2012 23:00:56", "content": "

The solution format could be ID with PREDICTEDAMOUNT2, or just PREDICTEDAMOUNT2 in the right order.

\r\n

Group is a 1-1 mapping with "jobid". I needed an integer for the back end scoring system, but you could ignore that and just use jobid.

\r\n

ID is a random number for each row of the test set. Shouldn't matter.

", "rawMarkdown": "", "votes": null }, { "id": "12895", "postDate": "08/03/2012 06:19:29", "content": "

[quote=DavidC;12890]

\r\n

The solution format could be ID with PREDICTEDAMOUNT2, or just PREDICTEDAMOUNT2 in the right order.

\r\n

Group is a 1-1 mapping with "jobid". I needed an integer for the back end scoring system, but you could ignore that and just use jobid.

\r\n

ID is a random number for each row of the test set. Shouldn't matter.

\r\n

[/quote]

\r\n

 

\r\n

Hi,

\r\n

Are the submissions to be sorted by GROUP and then by Predicted AMOUNT2?

\r\n

 

\r\n

I made my first submission with <ID>, <Predicted Amount2> and got AAT75=66339 and the following

\r\n

INFO: Assuming that column 1 with header value 'ID' maps to the required expected column 'amount2' (Line 1, Column 2)

\r\n

 

\r\n

Then I made a second submission with just the <Predicted Amount2> column {note this is sorted by ID} and got AAT75 = 0.63781 and the following message

\r\n

INFO: Assuming that column 1 with header value 'DONATED_Pred' maps to the required expected column 'amount2' (Line 1, Column 2).

", "rawMarkdown": "", "votes": null }, { "id": "12899", "postDate": "08/03/2012 16:37:23", "content": "

If it's just one column, then it needs to be sorted in the same way as the test data set (by "id").

\r\n

If you include the ID, you should make sure the header names match what the system expects ("id" and "amount2").

", "rawMarkdown": "", "votes": null } ], "comments": [ { "id": 12890, "author_name": "dchudz", "author_url": "", "post_date": "08/02/2012 23:00:56", "content": "

The solution format could be ID with PREDICTEDAMOUNT2, or just PREDICTEDAMOUNT2 in the right order.

\r\n

Group is a 1-1 mapping with "jobid". I needed an integer for the back end scoring system, but you could ignore that and just use jobid.

\r\n

ID is a random number for each row of the test set. Shouldn't matter.

", "votes": null, "replies": [] }, { "id": 12895, "author_name": "sashikanthdareddy", "author_url": "", "post_date": "08/03/2012 06:19:29", "content": "

[quote=DavidC;12890]

\r\n

The solution format could be ID with PREDICTEDAMOUNT2, or just PREDICTEDAMOUNT2 in the right order.

\r\n

Group is a 1-1 mapping with "jobid". I needed an integer for the back end scoring system, but you could ignore that and just use jobid.

\r\n

ID is a random number for each row of the test set. Shouldn't matter.

\r\n

[/quote]

\r\n

 

\r\n

Hi,

\r\n

Are the submissions to be sorted by GROUP and then by Predicted AMOUNT2?

\r\n

 

\r\n

I made my first submission with <ID>, <Predicted Amount2> and got AAT75=66339 and the following

\r\n

INFO: Assuming that column 1 with header value 'ID' maps to the required expected column 'amount2' (Line 1, Column 2)

\r\n

 

\r\n

Then I made a second submission with just the <Predicted Amount2> column {note this is sorted by ID} and got AAT75 = 0.63781 and the following message

\r\n

INFO: Assuming that column 1 with header value 'DONATED_Pred' maps to the required expected column 'amount2' (Line 1, Column 2).

", "votes": null, "replies": [] }, { "id": 12899, "author_name": "dchudz", "author_url": "", "post_date": "08/03/2012 16:37:23", "content": "

If it's just one column, then it needs to be sorted in the same way as the test data set (by "id").

\r\n

If you include the ID, you should make sure the header names match what the system expects ("id" and "amount2").

", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "12888": "", "12890": "", "12895": "", "12899": "" }, "source": "meta" }