Скачать или смотреть Underfitting Models with an Example || Lesson 39 || Machine Learning || Learning Monkey ||

#machinelearning#learningmonkey

In this class, we discuss Underfitting Models with an Example.
For understanding underfitting models we take an example of Linear Regression.

Here we take an example dataset having two columns.

We can plot this data in a two-dimensional coordinate space.

As we discussed machine learning models should identify the generalization of the data.

By taking a look at the data can we guess which machine learning model gives the best generalization on the data.

The data looks like a parabola.

The second-degree polynomial equation can generalize the data best.

By changing the a,b,c values in the equation we find the best parabola that fits our data.

Suppose if we apply our first base model on this data.

The base model is a Linear Regression.

It will identify a line that passing through the data and having a minimum loss.

The line identified is not at generalizing the data.

The models that are not concerned about the generalization of data we call it underfitting models.

We call under fitted models as high biased models.

By checking the above line what we can observe?

we use that line for predicting future data.

Let's check that the predicted value at five hundred is seventy-eight.

The predicted value at six hundred is also seventy-eight.

At all the x values we getting the same value seventy-eight.

This means our model is biased to one value.

That's the reason under fitted models are high biased models.

One important point we have to observe here.

According to data, our model may go underfit or overfit.

We have to identify the trade-off between overfit and underfit.

Under fitted models will have high training loss and high testing loss.

By using these values we identify whether our model is underfitting or not.

Link for playlists:    / @wisdomerscse  
Link for our website: https://learningmonkey.in

Follow us on Facebook @   / learningmonkey  

Follow us on Instagram @   / learningmonkey1  

Follow us on Twitter @   / _learningmonkey  

Mail us @ [email protected]