Comparison of regression
Evaluate Bayesian Linear Regression, Neural Net. Regression, Boosted Decision Tree Regression, Linear Regression, and Decision Forest Regr.
1. Data Boston upload from library(MASS)
2. Split data 50% train data and 50% test data.
3. Train data use medv as label
4. Model
Bayesian Linear Regression
Neural Network Regresiion
Boosted Decision Tree Regression
Linear Regression
Decision Forest Regression
5. Score Model
6. Evaluate Model
7. Execute R Script
For Bayesian Linear Regression upload R script:
>dataset <- maml.mapInputPort(1)
>data.set <- data.frame(Algorithm="Bayesian Linear Regression")
>data.set <- cbind(data.set, dataset[2:6])
> maml.mapOutputPort("data.set");
For Neural Net. Regression, Boosted Decision Tree Regression, Linear Regression, the R script:
>dataset <- maml.mapInputPort(1)
>data.set <- data.frame(Algorithm="Neural Net.work Regression")
>data.set <- cbind(data.set, dataset[1:5])
> maml.mapOutputPort("data.set");
8. Add Row, connected to all model.
9. Run, visualise the result.
10. Conclusion, model Boosted Decision Tree Regression is the best model with RMSE 2.790232 and
Coefficient of Determination 0.898652.
