The project worked on mining the previous airfare data and developing data modelling technique to predict the price variation over time. The purpose of this study was to better analyse the features that affect airfare, develop and tune models to predict the airfare well in advance and come with a easy to use, intuitive web application. ANN was used to accomplish this task.
The architecture used was as follows;
10 inputs, 32 neurons in the first hidden layer, 16 neurons in the second hidden layer, 1 output neuron. The final neuron gives the value of the predicted airfare.