Train a Linear Regression Model and Fit the Model

Quick reference

• 00:05 Okay, so we've imported our data.
• 00:08 We've created a data frame of that data, we've created our x and our y here and
• 00:12 our x is are column headers, and
• 00:14 our y is the price data of our actual housing stuff in Boston.
• 00:18 And now we've defined what parts of our model,
• 00:21 what parts of our data we want to train against and what we want to test against.
• 00:26 Now we need to actually train our model, our linear regression model.
• 00:29 So earlier we imported our model that we want to
• 00:32 use sklearn.linear_model is going to be linear regression.
• 00:35 So now we just need to create our linear regression instance or
• 00:39 our linear regression object.
• 00:41 And then just sort of run it, so we could do this very simply,
• 00:45 I'm just going to create a linear regression object.
• 00:48 And I'm just going to call it lin, you can call it anything you want.
• 00:52 And this is just going to be our linear regression instance, right?
• 01:00 So now we can type lin., and
• 01:03 then Shift Tab to learn all kinds of stuff about this.
• 01:07 So linear regression fits a linear model with coefficients
• 01:11 to minimize the residual sum of squared between the observed targets.
• 01:14 Remember I think of scatter plot, of the distance between the points in our scatter
• 01:18 plot and the line that goes through it, right?
• 01:20 fit_intercept is normalize, and if you want to come through here and
• 01:24 just read all of this stuff, it's kind of interesting stuff.
• 01:28 We're going to go through a lot of these things later on after we run this.
• 01:32 And there's even actually some examples that you can go through in here.
• 01:37 Very interesting.
• 01:39 See, we've done this earlier.
• 01:41 Very cool.
• 01:43 So, now that we've created an instance of our linear regression model,
• 01:48 we need to actually just train our data and fit the data.
• 01:51 And to do this, it's incredibly easy.
• 01:53 We just call our lin instance here and we say .fit.
• 01:57 And now if we shift enter, we see it go on and fit the linear model.
• 02:00 If we expand this, it needs parameters.
• 02:03 It needs an x and a y.
• 02:05 And remember, the x is our training data and the y is a target values,
• 02:10 in our case, it's going to be prices.
• 02:12 And you could put some sample way but we don't really care about that right now.
• 02:15 So we want to fit based on not our actual data but
• 02:20 our trained data from when we define X_train, X_test, y_train and
• 02:24 y_test, right, and we split our data between training data and testing data.
• 02:29 So here we need X_train, and we need y_train.
• 02:36 Now Shift Enter to run this and we see this linear regression copy_X=True,
• 02:41 fit_intercept=True, n_jobs=None, normalize=False.
• 02:46 This basically tells us that we're done,
• 02:49 it tells us that our linear regression model has been trained.
• 02:53 And if you're interested in what all of this stuff is,
• 02:56 we can go X_train and just kind of look at, see what these things are.
• 03:00 Oops, train, right?
• 03:03 So it's 303 rows of our data.
• 03:05 Remember our data was 500 and something but
• 03:08 40% of that goes to test size with leftover is 303 of the actual data.
• 03:14 We can also look at y_train.
• 03:17 And we're going to see something similar.
• 03:18 Remember, those are prices.
• 03:20 And those are 303 training data points,
• 03:23 the rest have been turned into testing data.
• 03:26 So our linear regression has been fit onto our training data.
• 03:30 And we're pretty much ready to go.
• 03:32 Now we can start to make predictions.
• 03:33 This is also generated now coefficients and intercepts and
• 03:36 we'll talk about coefficients and intercepts in the next video.

Lesson notes are only available for subscribers.