'Which is best model for regression if all independent variables are categorical and ordinal in nature?
I have dataset with 4000+ rows and 150+ features ,all of which are categorical. Though I'm getting a very less RMSE with Random Forest Regressor, when I plot actual vs predicted scatterplot, the values seem to be level-wise and widely scattered, the residuals vary from -15 t0 30. Let's have a look at the plot -
actual vs predicted scatterplot
Is it a right measure to evaluate performance of model through scatterplot ? Or how should I know the model is good or not for these complex relationship.
How does XGBoostRegressor would fit in this case ?
Sources
This article follows the attribution requirements of Stack Overflow and is licensed under CC BY-SA 3.0.
Source: Stack Overflow
| Solution | Source |
|---|