Tools
Tools
Pardeshi, Aaditya (2023) Hybrid Machine Learning Model for Auto Scaling using CPU Utilization. Masters thesis, Dublin, National College of Ireland.
Preview |
PDF (Master of Science)
Download (4MB) | Preview |
Preview |
PDF (Configuration Manual)
Download (3MB) | Preview |
Abstract
This project discusses an essential requirement in today’s world, that is by utilising strong machine learning models to increase auto-scaling capabilities, ensuring that cloud architecture readily aligns with changeable workloads. Inspired by the need for adaptable and cost-efficient serverless computing, the project introduces and evaluates four very known and important models such as Linear Regression model, Cat Boost Regressor model, Random Forest Regressor model and also a hybrid model that is a Voting Regressor ensemble. Dataset from Materna that contains CPU utilization metrics from VM’s is used. The implementation is conducted on AWS SageMaker. By using feature engineering and data preprocessing, the research investigates the effect of each model on system performance. Evaluation metrics like R-squared score, MSE that is Mean Squared Error, as well as Root Mean Squared Error (RMSE) is used for the evaluation. With an R2 score of 0.664, the proposed hybrid model’s performance was determined to be optimal and more efficient as compared with other individual models.
| Item Type: | Thesis (Masters) |
|---|---|
| Supervisors: | Name Email Sahni, Vikas UNSPECIFIED |
| Subjects: | Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Cloud computing Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning |
| Divisions: | School of Computing > Master of Science in Cloud Computing |
| Depositing User: | Ciara O'Brien |
| Date Deposited: | 09 Apr 2025 14:34 |
| Last Modified: | 09 Apr 2025 14:34 |
| URI: | https://norma.ncirl.ie/id/eprint/7399 |
Actions (login required)
 |
View Item |