Altair RISE
A program to recognize and reward our most engaged community members
Nominate Yourself Now!
Home
Discussions
Community Q&A
help me find a prediction of house price
tamararidwan
hi, i'm new in rapidminer and a student, i have a project to find a prediction of house price from the data i attach here. can anyone please help me to find the accuracy of the house price? since i tried so hard but always meet with rmse.
it would be great if anyone can help me.
thanks
Find more posts tagged with
AI Studio
Excel
Getting Started
Tutorials + Training
Predictions + Scoring
Algorithms
Accepted answers
lionelderkrikor
Hi
@tamararidwan
,
The accuracy is a performance metric dedicated to
classification
tasks.
Prediction of house price, like your project, is a
regression
task (target variable is
continuous
). So RapidMiner propose RMSE as performance metric.
However if you want a percentage as performance metrics for your regression task, you have to use (inside your
Cross Validation
operator) the
Performance (Regression)
operator and check
relative error
.
Hope this helps,
Regards,
Lionel
varunm1
Hello
@tamararidwan
@lionelderkrikor
already gave some great inputs, one thing to add is
squared correlation
(R squared, Range between 0 to 1 in rapidminer). The model is better if its closer to 1, most of the time above 0.6 is acceptable and it also depends on the problem domain.
All comments
robin
Can you post your process? Shout if you need help on how to post the code.
tamararidwan
@robin
is this process right? the result is rmse though, i want the result in accuracy percents
tamararidwan
@robin
can you please help?
lionelderkrikor
Hi
@tamararidwan
,
The accuracy is a performance metric dedicated to
classification
tasks.
Prediction of house price, like your project, is a
regression
task (target variable is
continuous
). So RapidMiner propose RMSE as performance metric.
However if you want a percentage as performance metrics for your regression task, you have to use (inside your
Cross Validation
operator) the
Performance (Regression)
operator and check
relative error
.
Hope this helps,
Regards,
Lionel
varunm1
Hello
@tamararidwan
@lionelderkrikor
already gave some great inputs, one thing to add is
squared correlation
(R squared, Range between 0 to 1 in rapidminer). The model is better if its closer to 1, most of the time above 0.6 is acceptable and it also depends on the problem domain.
tamararidwan
@lionelderkrikor
@varunm1
hi, i already tried what you said to me, but the result still rmse. can you help me?
thanks.
Quick Links
All Categories
Recent Discussions
Activity
Unanswered
日本語 (Japanese)
한국어(Korean)
Groups
