Paper Key : IRJ************732
Author: Pavan Kumar Kunisetty,Yerramsetti Sai Sindhu,K Vijay Kumar,G Mahesh Challari
Date Published: 01 Feb 2025
Abstract
This research investigates the symbiotic relationship between data quality, data quantity, and automated machine learning (AutoML), which has become indispensable as machine learning permeates diverse domains. The study aims to elucidate how bolstering data quality and optimizing data quantity influence the efficacy of AutoML pipelines. Through a comprehensive exploration of data quality assessment methods, data augmentation techniques, and their impact on AutoML outcomes, the research introduces a novel framework that seamlessly integrates data preprocessing, quality evaluation, quantity enhancement, and AutoML model selection. This framework not only offers practical guidance to enhance the efficiency of machine learning workflows but also unveils the intricate balance between data quality and quantity, showcasing scenarios where emphasizing one facet could yield superior results. Empirical findings underscore the proposed framework's potential to heighten predictive performance and generalization across tasks, bridging the gap between data quality, quantity, and automated machine learning. This contribution advances the AutoML domain, underscoring the holistic data preparation and model generation approach's significance, providing actionable strategies to augment the overall proficiency of automated machine learning systems sought after by organizations reliant on machine learning-driven decisions.
DOI Requested