My Math Forum  

Go Back   My Math Forum > Math Forums > Math

Math General Math Forum - For general math related discussion and news


Reply
 
LinkBack Thread Tools Display Modes
March 19th, 2017, 06:27 AM   #1
Newbie
 
Joined: Mar 2017
From: London

Posts: 3
Thanks: 0

Rating computing problem

Hello.

I am working on a formula for a rating computation out of number of elements in 5 categories. The general idea is as follows:

The possible number of elements for each category is different, and different should be the weight of influence for each of categories for the final outcome. In 4 categories the elements should have positive influence, and in on - negative.

Name / allowed range of number of elements / weight of element influence / influence + or -


Cat 1 / 0-20 / 1,5 / +
Cat 2 / 0-20 / 1 / +
Cat 3 / 0-5 / 3 / +
Cat 4 / 0-5 / 5 / -
Cat 5 / 0-20 / 0,5 / +


Now: the assumption is that final rating was in range 1-5.

If in data set will be:

- only 5 elements in Cat 4, then final rating should be 1.
- there will be no elements in any category, then final rating should be 3.
- there will be only 5 elements in category 4 and 5 in category 3, then final rating should be 4.
- only 5 elements in category 3, then final rating should be 4.5.
- all elements will be in any categories, but in category 4 there will be no element, then final rating should be 5.

I tried to use an online tool for Multiple Polynomial Regression:

Regression Tools - Online Multiple Polynomial Regression

But using the data below, the errors are very high (see attached image). What is the problem here, are there any inconsistencies in the data?

0 0 0 0 0 3
20 0 0 0 0 4
0 20 0 0 0 4
0 0 5 0 0 4.5
0 0 0 5 0 1
0 0 0 0 20 4
20 20 0 0 0 4.3
20 0 5 0 0 4.6
20 0 0 5 0 2
20 0 0 0 20 4.2
0 20 5 0 0 4.8
0 20 0 5 0 2.5
0 20 0 0 20 4.3
0 0 5 5 0 4
0 0 5 0 20 4.8
0 0 0 5 20 2
20 20 5 0 0 4.9
20 20 0 5 0 3
20 20 0 0 20 4.5
0 20 5 5 0 4.3
20 0 5 0 20 4.8
20 0 0 5 20 3
0 20 5 0 20 4.8
0 20 0 5 20 3.5
0 0 5 5 20 4.3
20 20 5 5 0 4.4
20 20 5 0 20 5
20 20 0 5 20 3.5
20 0 5 5 20 4.6
0 20 5 5 20 4.4
20 20 5 5 20 4.5
10 10 5 5 10 4


Last edited by skipjack; March 19th, 2017 at 04:05 PM.
Martin Cot is offline  
 
March 19th, 2017, 04:05 PM   #2
Global Moderator
 
Joined: Dec 2006

Posts: 17,728
Thanks: 1359

The image isn't viewable.
skipjack is online now  
March 20th, 2017, 12:48 AM   #3
Newbie
 
Joined: Mar 2017
From: London

Posts: 3
Thanks: 0

https://www.photobox.co.uk/my/photo/...id=21634574093

Last edited by skipjack; March 20th, 2017 at 02:23 PM.
Martin Cot is offline  
March 20th, 2017, 02:25 PM   #4
Global Moderator
 
Joined: Dec 2006

Posts: 17,728
Thanks: 1359

The image link is incorrect or the image is private and hence not viewable.
skipjack is online now  
March 21st, 2017, 04:40 AM   #5
Newbie
 
Joined: Mar 2017
From: London

Posts: 3
Thanks: 0

I am really sorry. Indeed, the image was viewable only from my machine.

I uploaded it to a different service and used TOR to check, it should be well readable now.


https://s10.postimg.org/vxmv1hlqx/table_new.jpg


Martin Cot is offline  
Reply

  My Math Forum > Math Forums > Math

Tags
computing, problem, rating, regression, statistics



Thread Tools
Display Modes


Similar Threads
Thread Thread Starter Forum Replies Last Post
Understanding the IMDb weighted rating function petervk Math 2 November 20th, 2015 06:08 AM
Rating mathematicians ferz Math 14 June 2nd, 2015 01:16 AM
Need to build a fair rating system. rohit_static Advanced Statistics 1 May 30th, 2012 06:21 AM
a Numerical Computing problem (please help me!) je.nickp Applied Math 2 October 24th, 2007 09:39 AM
Computing higher degree derivatives: hard problem julien Real Analysis 0 May 3rd, 2007 01:56 AM





Copyright © 2017 My Math Forum. All rights reserved.