ICP 5
In this ICP we studied another module of Machine Learning which is Linear Regression. We studied different types of Linear Regression. One is the Simple Linear Regression and the other is Multi Regression. In simple linear regression there is only one feature whereas in multi regression there are more than one features.
The tasks are as follows.
- Delete outliers for the GarageArea column.
- Implement Linear Regression on the wine quality dataset and evaluation of the model.
In this task we first evaluated the outliers which can be seen using the scatter plot.
There are different ways to remove the outliers. For this data set I have used the method of Inner Quartile Range (IQR) method. The image below shows the scatter plot after the removal of outliers.
In this task we had to perform linear regression on the wine quality data set. In order to get the correlation I applied the function of generating heat map in order to visualize the correlation matrix.
Removal of null values can be shown in the following screenshot.
The top 3 correlation features are
Bottom 3 correlation features are
The final result of regression after training the models is given below.
Evaluation using RMSE and R2 score is shown below.
