Enhancing Airlines Delay Prediction by Implementing Classification Based Deep Learning Algorithms

Abstract

Technology is evolving in a rapid pace with its numerous discoveries, and nowadays, the rate is more than ever before. Data Analytics has become a knowledge and a tool which significantly contributing relentlessly to majority of the discoveries since this can fetch insights to reduce man-machine interactions. Prediction is an integral part of data analytics that provides meaningful information from historical data to support decisions. Machine learning and deep learning is the core of any predictive analytics where both have their own strengths and weakness. Aviation industry around the world are facing severe problems by the flight delays caused by several factors. In order to achieve its target to provide a hassle-free journey, Aviation Industry is continuously researching to reduce flight delays. This research will focus mainly to predict airlines flight delays by analyzing flight data, especially, for the domestic Airlines those moves around the United States of America. Data science methodology has been implemented in order to fetch the end prediction. In order to transform the high dimension data into a low dimension Principal component analysis is used. Deep learning algorithms, widely popular and state-of-the-art prediction technology, are implemented in the prediction modeling phase. By empirical observation, the research can come to a conclusion that by following the data science methodology better performance could be unlocked to help the aviation industry.