Optimizing Machine Learning with PCA: A Study on Efficiency and Model Performance
Keywords: Principal Component Analysis, Machine Learning, Dimensionality Reduction, Feature Extraction, Computational Efficiency
ABSTRACT
Principal Component Analysis (PCA) remains a core technique in the machine learning toolkit, particularly for simplifying high-dimensional data. This study examines how PCA can be useful for dimensionality reduction, data compression and in improvement of model performance. We perform an original experiment using a synthetic dataset and real dataset to evaluate impact of PCA on classification accuracy and training time using multiple machines learning algorithms. The results indicate that PCA significantly reduces computational power while maintaining and improving predictive performance of big dimensional data. We also address inherent limitations of PCA, such as its linearity assumptions, and suggest paths for extending its capabilities. Overall, our analysis highlights PCA’s enduring relevance and the need for further innovation in dimensionality reduction strategies. This study emphasize PCA’s critical role in modern machine learning pipelines as well as its potential for further improvement in handling complex datasets.
"Optimizing Machine Learning with PCA: A Study on Efficiency and Model Performance", IJSDR - International Journal of Scientific Development and Research (www.IJSDR.org), ISSN:2455-2631, Vol.10, Issue 7, page no.c415-c418, July-2025, Available :https://ijsdr.org/papers/IJSDR2507264.pdf
Volume 10
Issue 7,
July-2025
Pages : c415-c418
Paper Reg. ID: IJSDR_304366
Published Paper Id: IJSDR2507264
Downloads: 000548
Research Area: Science and Technology
Country: Jalgaon, Maharashtra, India
DOI: https://doi.org/10.56975/ijsdr.v10i7.304366
ISSN: 2455-2631 | IMPACT FACTOR: 9.15 Calculated By Google Scholar | ESTD YEAR: 2016
An International Scholarly Open Access Journal, Peer-Reviewed, Refereed Journal Impact Factor 9.15 Calculate by Google Scholar and Semantic Scholar | AI-Powered Research Tool, Multidisciplinary, Monthly, Multilanguage Journal Indexing in All Major Database & Metadata, Citation Generator
Publisher: IJSDR(IJ Publication) Janvi Wave
