Training Loss and Testing Loss || Lesson 37 || Machine Learning || Learning Monkey ||

#machinelearning#learningmonkey

In this class, we discuss Training Loss and Testing Loss.
We take an example and understand training loss and testing loss.

The example we consider here is the tip amount prediction data set.

In this, we have two columns.

Bill amount.
Tip amount.

Based on the bill amount tip amount will be predicted.

We consider here ten data points.

The first seven data points are taken as training data.

The last three points are taken as testing data.

Training data is given to a machine learning model as input.

The machine learning model which we consider here is a linear regression model.

What our model will do?

Take training data as input and generate an equation of a line that passes through the data.

This line will have a minimum loss.

Now we use this line equation for predicting future data points, ie testing data.

Lets assume our model identified a line y = 0.1x + 2.

Take the testing data and substitute x value in the above function.

we get the predicted values.

In the data set, we had the actual tip amount given by the customer.

Loss is given as an actual tip amount minus predicted tip amount. and squared.

We square the value for eliminating negative signs.

Like this calculate the loss for all the values in the test data and divide by the number of data points.

The value we call testing loss.

If we do the same process on training data we get training loss.

we use this training loss and testing loss in all the machine learning model.

Based on the training and testing loss values we decide how good our machine learning model is.

The loss calculation is already discussed in our linear regression class.



Link for our website: https://learningmonkey.in

Follow us on Facebook @   / learningmonkey  

Follow us on Instagram @   / learningmonkey1  

Follow us on Twitter @   / _learningmonkey  

Mail us @ learningmonkey01@gmail.com