Airline-Customer-Service-Analysis

Table of Contents

1. Tools used
2. Overview
3. Goal
4. Technical Aspect
5. Data Flow
6. Exploratory Data Analysis results
7. Evaluation of Machine Learning models
8. Conclusion/Discussion

1. Tools used

2. Overview

This project contains Exploratory Data Analysis (EDA) and evaluation metrics for different Machine Learning models in python.

3. Goal

The goal for this project was to use data preprocessing and exploratory data analysis to find reasons for customer dissatisfaction. By analyzing the dataset and using different machine learning techniques to model passenger satisfaction. Recommendations were drawn from the insights generated through EDA which helped with improving customers satisfaction for the airline.

4. Technical Aspect

The data analysis for this project was dividied into:

• Data cleaning
• Data wrangling
• Exploratory Data Analysis
• Machine Learning
• Report writing/Presentation

5. Data Flow

6. Exploratory Data Analysis results

7. Evaluation of Machine Learning Models

* Random Forest result

* K Nearest Neighbors result

8) Conclusion/Discussion

• Random Forest algorithm had the highest accuracy (0.94) compared to KNN's (0.83).
• Majority of the customers (56.7%) were dissatisfied with the services provided by the airline.
• Adults were the most frequent users with more than 60000 count.
• Inflight wifi service, Ease of online booking, and online boarding were some of the factors that passengers rated with lower ratings 3 or below on a scale of 1 to 5 with 5 being the strongly liked.