Sunday, March 17, 2013

Regression Analysis ?



Regression Analysis
In many of the recent blog posts, along with several discussions in class, we have touched upon the topic of regression analysis.  In this post, I will elaborate on the basics of regression analysis
Regression analysis is a method of statistical analysis which estimates the relationship among variables. This method focuses in the relationship between a dependent variable (y) and one or more independent variables (the variables that affect the thing you want to calculate). Usually, this method estimates something called “conditional expectation,” which refers to the average value of the dependent variable (the variable you want to calculate) when the other variables are fixed.
Regression analysis is mostly used for prediction and forecasting. When discussing many of the relationships in class, we mentioned that we could make equations to predict certain outcomes. A regression model is presented as an equation, which solves for a dependent variable (y) given the known values of the X variables.
Y= K + AX1+BX2
·          Being A and B coefficients
·          K is a constant ( similar to a y-intercept)
·          X1 and X2 being variables

When we predict within the range of values that we have in our data set, we are interpolating our data. When we are predicting values outside of our range, we are extrapolating. 
Usually, it is easier to trust the interpolation of our data as we have something to back it up. On the other hand, the extrapolation can be misguiding. When we are predicting something outside of our range, we assume that the conditions are the same and this might not be the case. A rule of thumb is that the farther away we extrapolate our data, the more likely we are to make a mistake in our predictions.

2 comments:

  1. I remember you mentioning this in class and I was curious about what you meant by this. Thanks for explaining this to us in a blog post. I didn't really understand the different between interpolation and extrapolation, and the impact it had on data.

    ReplyDelete
  2. A picture would easily clarify the difference between interpolation and extrapolation (but I can't put pictures in comments). Xavi, you could also explain this in terms of the example that brought it up in class in the first place.

    ReplyDelete

Note: Only a member of this blog may post a comment.