What tools can we use to solve LR equations to get unknown parameter vector?

olap

New Member
#1
Greetings,

Apologies if my explanations are unclear, I am new to data mining coming from business intelligence development area.

Let's say we have the total daily revenue R that we get selling a few hundred different products (p). We know the daily volumes sold (v) and need to use LR to estimate each product's revenue per unit (rpu). The simplified daily revenue equation for day will look like:

R=v1*rpu1+v2*rpu2+...+vn*rpun

Obviously, we will have many days with their associated R

To summarize. the regressands R and regressors v are known. We need to get parameters rpu. The data to be mined can be stored in a SSAS cube or sql table(s) .

I would like to get an advice if it's possible to use SSAS data mining to tackle the task, which model to use and if possible, what keys to assign etc OR what other applications could be utilized. Thanks!
 

Dason

Ambassador to the humans
#2
It sounds like you're just looking to fit a regression? Any stats package can do that for you. Excel should even be able to do it.
 

olap

New Member
#3
Thanks Dason. I am looking into some online Excel tutorials. Say we just take a single day with R. In Input Y Range option do I just select one cell containing R? and what do I do for multiple days each containing R and v. Thanks!
 

Dason

Ambassador to the humans
#4
I don't know how to do it with Excel. I avoid Excel as much as humanly possible. But I know most people are more comfortable with Excel than other programs and if you don't have any other stats software it's hard to recommend R (the stats programming language - not the R in your problem) since it has a steep learning curve (but if you plan on doing serious stats it's worth it to learn).