Now, let's start the sixth step of our pipeline i,e Fine tune the model.
After trying out several models, we shortlist the best ones. In this chapter, we have only tried three models, but it is advised to try out more models so we can get a list of promising models.
We not only select a single best model but a list of promising models because it may be possible that after fine tuning the models, some other model starts performing better than the current best model.
We'll fine tune RandomForestRegressor because other two models (i.e., LinearRegression and DecisionTreeRegressor) performed very badly. If we would have tried some other models such as Support Vector Machine or Polynomial Regression we would have got some more models to fine-tune. But our goal is only to give you an understanding of the flow of a machine learning project. But you are free to try other models too and shortlist the best ones.
While fine-tuning the model, we try out different combinations of its hyperparameters and select the combination which gives us the best performance. Different combinations may work out well for different datasets. We can do this manually but it is a tedious process. Hence we perform this task using GridSearchCV from sklearn.
Taking you to the next exercise in seconds...
Want to create exercises like this yourself? Click here.
No hints are availble for this assesment
Answer is not availble for this assesment
Loading comments...