Embedded Files
Feature Significance
Feature Significance
In order to understand how the parameters affect our model, we first look at how the parameters are related to each other.
From the plot, one can infer that the total fare amount is highly correlated to the trip distance and the tip amount is also correlated to the total fare amount.
This analysis gives an idea of which features play an important role in our classifier and can help in further tuning the model.
Training Results:
Training Results:
Frequency Distribution:
Frequency Distribution:
From the above frequency distribution, one can infer that a high level of predictions resulted in a category as `Medium`. This is backed by the confusion matrix shown below.
Confusion matrix
Confusion matrix
The results from confusion matrix also coincides with frequency plot and one calculate the accuracy of the model.
The accuracy is around 89.62% for training.
Predicted Results:
Predicted Results:
Confusion Matrix
Confusion Matrix
The results from confusion matrix coincides with training and one calculate the accuracy of the model.
The accuracy is around 92.62% for training.
Page updated
Report abuse