cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
Check out the JMP® Marketplace featured Capability Explorer add-in
Choose Language Hide Translation Bar
jonessas
Level I

STATISTIC PROBLEM

Fit a KNN model to the Titanic Passengers dataset where "Survived" is the response and all other variables (except Name) are the predictor (X) variables. Set the validation to the validation column.  Set the K to 15, and the random seed to 125. Based on the results, what is the optimal K (enter the numerical answer, e.g., 3)

 

 

 

Titanic Passengers" JMP dataset

1 ACCEPTED SOLUTION

Accepted Solutions
txnelson
Super User

Re: STATISTIC PROBLEM

Is there a question you want to ask? Are you asking how to open the Titanic Passengers data table? Or how to access the K Nearest Neighbor Platform to run the analysis?
The data table can be accessed under the Help pull down menu
Help=>Sample Data Library
The Nearest Neighbor Platform is available under the Analyze pull down menu
Analyze=>Predictive Modeling=>K Nearest Neighbors
Jim

View solution in original post

2 REPLIES 2
Thierry_S
Super User

Re: STATISTIC PROBLEM

Hi,
Forgive me to ask but this looks like a homework assignment. If it is, it might be best to try to figure out by yourself through exploration of the KNN platform in JMP.
One point though, when you enter all predictors listed in the table, the KNN platform takes a long time to compute and also, some of the columns a very sparsely populated leaving most passenger with "empty" values. Hence, it might be valuable to experiment with simpler models that only include key variables such as Age, Sex, and Passenger Class.
Best,
TS
Thierry R. Sornasse
txnelson
Super User

Re: STATISTIC PROBLEM

Is there a question you want to ask? Are you asking how to open the Titanic Passengers data table? Or how to access the K Nearest Neighbor Platform to run the analysis?
The data table can be accessed under the Help pull down menu
Help=>Sample Data Library
The Nearest Neighbor Platform is available under the Analyze pull down menu
Analyze=>Predictive Modeling=>K Nearest Neighbors
Jim