Hello,
I am trying to score new data. I have a classification problem. One of the variables is the target. It is set to 1 if true and 0 if false. But how will i know from the score node output, how many have been classified into one class and another into another class? Isn't that the point of scoring?
Thanks
Dee
For your analysis, SAS Enterprise Miner generated
P_PATIENT_ALIVE_FLAG1: the probability that PATIENT_ALIVE_FLAG = 1
P_PATIENT_ALIVE_FLAG0: the probability that PATIENT_ALIVE_FLAG = 0
The assignment to an outcome is ultimately a modeling decision. In your previous note, you shared results which indicated that you had a target variable named
PATIENT_ALIVE_FLAG
which could be either 1 or 0 (alive or dead, presumably). In this case, SAS Enterprise Miner would create two prediction variables to store the probability of each outcome:
P_PATIENT_ALIVE_FLAG1 = the probability that PATIENT_ALIVE_FLAG=1 for the observation
P_PATIENT_ALIVE_FLAG0 = the probability that PATIENT_ALIVE_FLAG=0 for the observation
as well as an 'Into' variable which predicts which class is most likely based on the larger of the two values above, and this variable would be named (in this example)
I_PATIENT_ALIVE_FLAG = the most likely outcome based on the two probabilities described above
It is possible, however, that you might not want to predict PATIENT_ALIVE_FLAG = 0 unless there is a very high probability that this was the case (e.g. perhaps only when P_PATIENT_ALIVE_FLAG0 is greater than 0.9). In any case, the software can compute the probabilities but you need to decide what the threshold should be to determine if you predict PATIENT_ALIVE_FLAG=1 or PATIENT_ALIVE_FLAG=0. It is then easy to assign the observations to a category based on the cutoff you choose.
Hope this helps!
Doug
For your analysis, SAS Enterprise Miner generated
P_PATIENT_ALIVE_FLAG1: the probability that PATIENT_ALIVE_FLAG = 1
P_PATIENT_ALIVE_FLAG0: the probability that PATIENT_ALIVE_FLAG = 0
The assignment to an outcome is ultimately a modeling decision. In your previous note, you shared results which indicated that you had a target variable named
PATIENT_ALIVE_FLAG
which could be either 1 or 0 (alive or dead, presumably). In this case, SAS Enterprise Miner would create two prediction variables to store the probability of each outcome:
P_PATIENT_ALIVE_FLAG1 = the probability that PATIENT_ALIVE_FLAG=1 for the observation
P_PATIENT_ALIVE_FLAG0 = the probability that PATIENT_ALIVE_FLAG=0 for the observation
as well as an 'Into' variable which predicts which class is most likely based on the larger of the two values above, and this variable would be named (in this example)
I_PATIENT_ALIVE_FLAG = the most likely outcome based on the two probabilities described above
It is possible, however, that you might not want to predict PATIENT_ALIVE_FLAG = 0 unless there is a very high probability that this was the case (e.g. perhaps only when P_PATIENT_ALIVE_FLAG0 is greater than 0.9). In any case, the software can compute the probabilities but you need to decide what the threshold should be to determine if you predict PATIENT_ALIVE_FLAG=1 or PATIENT_ALIVE_FLAG=0. It is then easy to assign the observations to a category based on the cutoff you choose.
Hope this helps!
Doug
Don't miss out on SAS Innovate - Register now for the FREE Livestream!
Can't make it to Vegas? No problem! Watch our general sessions LIVE or on-demand starting April 17th. Hear from SAS execs, best-selling author Adam Grant, Hot Ones host Sean Evans, top tech journalist Kara Swisher, AI expert Cassie Kozyrkov, and the mind-blowing dance crew iLuminate! Plus, get access to over 20 breakout sessions.
Use this tutorial as a handy guide to weigh the pros and cons of these commonly used machine learning algorithms.
Find more tutorials on the SAS Users YouTube channel.