Using SAS data miner - Logistic regression

[Music] in this session I will show you how to use data miner frequent restaurants or not based on thousands of dollars variable that depicts whether that person frequents restaurants or not now how do we run this degradation the data is already saved as CSV file to input this data in and then the logistic regression and then I'll also see some minor you need the jnlp file earlier a few seconds to load [Music] take you let's yeah Enterprise miner running a superior project and this can be saved on the server let me just call it a gesticulation next all right can we have it team we need to create a new diagram a new diagram is Biscay is right so then we import the data in first import touken import file my computer browse and let's go to my folder I find my CSV file and open it alright let's go ahead and set up our variables income is input frequent restaurants is going to be our target and it is a binary like okay next step is to do the regression model and it should be under model drag and drop it on the canvas linked us to her and notice it you know it has already automatically selected logistic regression as the regression type and it's going to use the logit function that is fine with us and there's nothing else there's no changes to here just God and run it earlier I had run this logistic regression in Excel and got some results and the results was something like this the regression result was the logit equals negative nine point six three two plus coefficient point three six or eight times X and that's what we hope to find in sizes output also running all right we have the results now it's run so let's take a look at results the results will have some analysis screens as well as lots of different statistics but we are currently interest rate only in the intercept in the coefficient and let's try and find it okay let's scroll down here is output right at the very bottom I believe is where we have okay results of the regression equation there it is and also the maximum likelihood estimates that's what we do when we do logistic regression so intercept is negative nine point six three one six and the coefficient is point three six oh eight it's just the same as what we had seen earlier using Excel all right now how do we do a chart of this data and so we just close this for now and let's go back to the I would like to graph it so I will use Explorer of the plot the multi plot to allow me to create scatter plots and any other type of plot I want I want to see the s-shaped curve okay drag and drop this and what we do is you run it when we ran it it'll then show us a screen which is interactive where you can create charts on the fly alright I'll complete it now let's take a look at results the screen Digital darling okay all right multi plot let's see let me use a different graphics flow I should have chosen craft Explorer that was the one that would let me explore different graphs and okay it's loading I connect this to here I delete that and then I run it so drowning flow okay here we have it and let's take a look at the results of graphics power all right now this one has a little tool that lets you create plots so I'm gonna click on that and let's first create a chart between the income and frequents restaurant scatterplot can come on the x-axis frequents restaurants on the y-axis that's what it looks like let's do one more store between the predicted odds ratio and income income is x-axis [Music] so the here [Music] [Music] [Music] [Music] let me save this one too I'll print it as a PDF and let me go to the other window click on that and print it my PDF writer let me call this logistic regression list of three okay let's take a look at those files all right let's like the outcome at the logistic regression and this is original data and this is freaking stressed French I don't know what to write dependent all right now I can close this back here you project clock logs doesn't show you the result just information about the runs that we did all right if you wanted the actual digression [Music] [Music] thank you

Channel: Solomon Antony

Views: 930

Rating: 2.3333333 out of 5

Keywords:

Id: JeJTEc1AjLI

Length: 23min 21sec (1401 seconds)

Published: Wed Feb 08 2017