So, our RMSE for testing data comes to around 47000. It is not a very good model but our task never was to build a good model but only to understand the flow of building a machine learning project.
But, we can't be fully sure of the performance of our model by just an RMSE value before launching the system. So, we use several statistical measures to gain confidence in the performance of our model.
We then compare the performance of our model with the model currently in production. If our model is remarkably better than the current solution, then we proceed to the next step otherwise we optimize our model further.
We not only judge the model's performance on basis of error but also on other factors such as-
and many other factors. And, by combining these all factors by weighing them according to our needs, we evaluate the performance of our model.
Suppose our newly trained model's error is only 0.7% less than the model currently in prediction but it predicts the output in much less time. So, this will also be a remarkable improvement in the current solution.
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...