Airline Passenger Satisfaction Prediction

Keywords: Airline Passenger

Background

Airline passenger satisfaction is a crucial aspect of the aviation industry, influencing customer loyalty and brand reputation. In this study, the author investigates the use of Artificial Neural Networks (ANN) for predicting passenger satisfaction based on a dataset containing various attributes related to the airline experience. The target variable consists of binary classes, namely “neutral or dissatisfied” and “satisfied”. The dataset is preprocessed to handle missing values, the categorical variables were encoded using label encoding techniques. Remaining numerical input variables were rescaled using the techniques of standardization. The ANN model is designed with multiple layers, including an input layer, hidden layers with activation functions of ReLU, and an output layer with a sigmoid activation function. To mitigate the problem of overfitting, Batch Normalization was also done in the layers. A total of 5 layers were built including an input and output layer. The input activation function is ReLu, and the output activation function is sigmoid, as there are two classes in the target variable i.e., satisfaction. The optimizer used was Adam, and the loss which was applied in the current ANN model is binary cross entropy. After all these parameters applied, the accuracy of the model is 96 percent.

Aim & Objectives

Aim
To Predict the Airline Passenger Satisfaction Using DL Techniques
Objective

Research Methodology

ANN Technique used

Software & Tools

Jupyter, Google Collab

— Preview truncated. The complete document (full chapters, references, diagrams, and appendices) is shared with clients as part of project delivery. —