Downloads: 18 | Views: 128 | Weekly Hits: ⮙7 | Monthly Hits: ⮙7
Research Paper | Computer Science and Information Technology | United States of America | Volume 14 Issue 4, April 2025 | Popularity: 7.3 / 10
AI - Driven Airfare Prediction Models for Cost Optimization and Consumer Savings in the U. S.: Integrating ETL and Cybersecurity for Enhanced Data Processing and Protection
Shiva Kumar Vuppala
Abstract: Airlines compete fiercely on airfare, and the price is highly dynamic with the changing season of demand, fuel prices, and large economic conditions. Existing traditional pricing models tend not to respond easily to changes in the market, and so there exists a need for AI - driven, predictive - based approaches to help make the fare changes more accurate and efficient. The current study is an attempt to design an advanced AI - powered framework to predict airfare and implement dynamic pricing optimization using time series forecasting (ARIMA), machine learning (XGBoost) and Reinforcement Learning (DQN - RL). Furthermore, the essay also evaluates the effectiveness of using ETL (Extract, Transform, Load) pipelines on huge scales of real - time airfare data and security measures such as data integrity and privacy for the data itself. The novel aspect of this research is that, in contrast to conventional statistical models, the model treats ARIMA for long - term trends in fare forecast, XGBoost for short - term price fluctuation, and RL - DQN for real - time booking recommendation according to expected price change. The framework proposed in this work combines ETL automation to fetch, clean and normalize historical and real - time fare data for model retraining, which is done in real - time. Results show that ARIMA reaches 94.5% accuracy in long - term forecasting and XGBoost 90.2% accuracy for short - term fare prediction development. In dynamic pricing recommendations, the RL - DQN model outperforms both and gets 95.2% accuracy. In addition, the ETL pipelines run in an efficient manner processing large scales of real - time fare data with an average API response time of 120 - 130 milliseconds. This confirms that Airfare prediction by AI improves accuracy and helps save costs for travelers as well as airlines. Additionally, automated ETL pipelines improve the reliability of data, while strong cybersecurity measures guarantee the safe processing of the fare data.
Keywords: AI - Driven Pricing, Airfare Prediction, Machine Learning, ETL Processes, Cybersecurity, Predictive Analytics, Cost Optimization
Edition: Volume 14 Issue 4, April 2025
Pages: 230 - 242
DOI: https://www.doi.org/10.21275/SR25401070655
Please Disable the Pop-Up Blocker of Web Browser
Verification Code will appear in 2 Seconds ... Wait