Let’s make predictions on that training data, to see how well our model fit it during training.
NOTE: Training error measures how well your model fits the training data, but it does not measure how well your model generalizes to new data. In later exercises, you’ll explore how to split your data to evaluate your model’s ability to generalize.
Is this a good model? How would you judge how large this error is?
Mean Squared Error (MSE) can be hard to interpret, so we often look at Root Mean Squared Error (RMSE) instead. A nice property of RMSE is that it can be interpreted on the same scale as the original targets.
Let’s compare the RMSE to the difference of the min and max of our targets:
Our error spans nearly half the range of the target values. Can we do better?
This is the question that nags at every model developer. Let’s develop some basic strategies to reduce model error.
The first thing we can do is take a look at how well our predictions match our targets, in terms of overall summary statistics.
Okay, maybe this information is helpful. How does the mean value compare to the model’s RMSE? How about the various quantiles?
We can also visualize the data and the line we’ve learned. Recall that linear regression on a single feature can be drawn as a line mapping input x to output y.
First, we’ll get a uniform random sample of the data so we can make a readable scatter plot.
That’s it our first linear Regression model. Remember that I have posted images of code so that you should not copy and paste it just work our on your own.
We will be driving deep from next classes.
Link to my blog : https://machinelearningpower.home.blog.com
Machine Learning 
I have mixed this model with another one, Here in some places, there is California Dataset just remember that our dataset name is “Data_Frame”
LikeLike