how to found out prediction of a data house price?

tamararidwan

hi, so I'm new in this Rapidminer and i have some difficulties, it would be great if someone can help me
i wanted a prediction of a data house price per HousingMedianAge, but it seems like the result always root_mean_square_error. i'm not sure what was wrong since i'm new and a student as well. the data is in numerical

and i'm not really sure if the whole models i made was right or not


but it might be wrong since the result is always error, but i found a formula to found out the result of te prediction using rmse formula
 

best regard,
Tamara

tftemme

Hi @tamararidwan,

What you are looking at is the performance of your prediction model. The root mean squared error is a performance measure to describe the performance of a regression task (see https://en.wikipedia.org/wiki/Root-mean-square_deviation). The Cross validation operator trains your model of a subsample of your data and test its prediction on another subsample (for more details please check the operator documentation, RapidMiner Academy or most text ressources about data science/machine learning). 
Please connect the 'tes' (test) output port of the Cross Validation operator to the result, to see the actual predictions your model perform on the test set. Also you can double click into the Cross Validation operator, select the 'Performance' operator in the testing subprocess and add more performance measures (for example relative error, which is often more interpretable).

Hopes this helps
Fabian

tftemme

Hi @tamararidwan,

What you are looking at is the performance of your prediction model. The root mean squared error is a performance measure to describe the performance of a regression task (see https://en.wikipedia.org/wiki/Root-mean-square_deviation). The Cross validation operator trains your model of a subsample of your data and test its prediction on another subsample (for more details please check the operator documentation, RapidMiner Academy or most text ressources about data science/machine learning). 
Please connect the 'tes' (test) output port of the Cross Validation operator to the result, to see the actual predictions your model perform on the test set. Also you can double click into the Cross Validation operator, select the 'Performance' operator in the testing subprocess and add more performance measures (for example relative error, which is often more interpretable).

Hopes this helps
Fabian

tamararidwan

Hi @tftemme thanks for your help, i will try it out