Introduction

The objective of this exercise is to construct multiple models capable of accurately predicting the total price of taxi fares. To begin, a basic multiple regression model will be developed. Afterward, a random forest model will be created to see if we can do better. Finally, an Artificial Neural Network (ANN) will be implemented. ANNs are mathematical models designed to replicate the information-processing capabilities of the human brain. By utilizing these models, I aim to achieve robust and precise predictions of taxi fares.

The dataset comes from Kaggle.

The question I want to answer

Given that there are two passengers on a trip with a trip duration of 100 minutes, a distance traveled of 5 miles, a tip of $10, and miscellaneous fees of $5, what will be their total fare?

Data Preprocessing

Normalize data

The total fare column is now normally distributed. We can now take care of the trip duration column.

Linear Regression with Sklearn

Random Forest Regression

Artificial Neural Network

The Question and Answer

Given that there are two passengers on a trip with a trip duration of 100 minutes, a distance traveled of 5 miles, a tip of $10, and miscellaneous fees of $5, what will be their total fare?

Based on the model, I expect the total fare to be $54.25