Cloud service price prediction using Machine Learning Algorithm with API in the case of Amazon Web Services and Microsoft Azure
B Dhayanandan, R Rajeev · 2024
Due to the availability of highly scalable and reasonably priced infrastructure, software, and platform services, many enterprises are moving their on-premise artificial intelligence workloads to the cloud. The study focuses on the K-Nearest Neighbors (KNN) algorithm and uses a quantitative method to build and assess a predictive model for cloud service price prediction. Pre-processed historical pricing data from Microsoft Azure and AWS is gathered, together with variables that affect the cost of cloud services. In order to generate models, the KNN algorithm is used, and an API is made to integrate real-time price predictions. Evaluation measures (MAE, MSE, and RMSE) are employed. Because cloud pricing is unpredictable, the suggested method addresses this dynamic by improving accessibility and adaptability. Findings from exploratory data analysis and features of the API show how effective the system is. This contribution involves providing customised solutions for each customer while taking ethical and practical constraints into account. All things considered, the integrated approach is a major improvement in cloud service pricing prediction, allowing for “well-informed” budgetary and resource allocation decisions.