Consisting of accidental and systematic errors.

Concretely, let our training error and cross validation error be defined as in the previous videos, just say, the squared error, the average squared error as measured on the 20 sets or as measured on the cross validation set.

So one fundamental lesson to take away from that simple example is look at your residuals.

Here's what I mean.

This is where the concept of ordinary lease squar es regression comes in. And the idea is very simple.

And similarly, I'm then also going to define the cross validation set error when the test set error, as before to be the average sum of squared errors on the cross validation and the test sets.

So of course, the difference between the actual score Y, or observed score Y and predicted score Y is error.

They did much better than 2100. So the distance between.

So, now that we have defined the training validation or cross validation and test sets, we can also define the training error, cross validation error, and test error.

Whereas, your cross validation error, assuming that this say the squared error which we're trying to minimize.

Variance

Error diffusion

So B sub zero again is just the intercept for the slope of the line.

It goes from two forty down to two hundred.

He actually went to UlC where I use to teach, I think he actually took my stats course.

I can delete that dot because that's clearly a measurement error.

Monochrome error diffusion

Having done that, finally, what

So we can see the ratio of this to this.

And then j subscript cv is my cause validation error is pretty much what you'd expect.

If the prediction of our function f equals our target label, and the error is zero, then no update occurs.

So, if your algorithm is suffering from a bias problem, the training set error would be high and you may find that the cross validation error will also be high.

Concretely, for the high bias case, that is, the case of under fitting, what we find is that both the cross validation error and the training error are going to be high.

Give or take a few years

J C V and J

Get our predicted score on y, it's a function of the regression constant plus some predictor times its' slope.

That's no problem, as most of you know, because I can just delete that dot.

So, I'm gonna introduce a new formula and a new concept here, and one that will be, essential as we move forward in inferential statistics.

If I go to my output for 100 steps, I find that the y value is between 0.7 and 0.9.

If your hypothesis basically classified the example y correctly.

You see a big prediction error for Curtis Granderson. Same for Eduardo Nunez.

It's usually the same thing as the j of theta that we'll be writing so far, this is just a training set error you guys measure on your training set.

Values given over accidental errors.

And of course, this is the definition of the test set error if we are using linear regression and using the squared error metric.

Yeah. I would agree with that.

Even though its value changed by a whole order of magnitude.

No Accounting

Pressure

Measure

Still a pretty big prediction error, but not as bad. Nunez is way down here.

The answer is this.

learning algorithm and it is not performing as well as your are hoping, so your cross validation set error or your test set error is high.

Right, that's sampling error. And how much sampling error we expect on average is standard error.

As N gets really, really large.

If you draw a curve like this, your quadratic error becomes 2.