Authors :
Prathamesh Sunil Patil
Volume/Issue :
Volume 8 - 2023, Issue 11 - November
Google Scholar :
https://tinyurl.com/mrxkk497
Scribd :
https://tinyurl.com/yrubney4
DOI :
https://doi.org/10.5281/zenodo.10203468
Abstract :
In the realm of machine learning, the
dimensions of data have long been a double-edged sword
– offering both promise and peril to its practitioners.
Through a comprehensive study data available in
literatures, real-world applications and practical
experiments, we elucidate the formidable curse of
dimensionality and its adverse effects on model
generalization, computational resources and
interpretability. Furthermore, we delve into the arsenal
of dimensionality reduction techniques and feature
selection strategies, revealing the power of transforming
the data into actionable insights. This paper
demonstrates the tangible benefits of effectively
managing dimensions in machine learning, providing
practitioners with invaluable insights to harness the true
potential of their data. To validate the efficacy and
reliability of our proposed methodology, I conducted a
case study using a simple and informative dataset,
specifically focusing on Iris dataset.
Keywords :
Machine Learning, Informative and Simple Dataset, Dimensionality Reduction, PCA, LDA, Dashboards.
In the realm of machine learning, the
dimensions of data have long been a double-edged sword
– offering both promise and peril to its practitioners.
Through a comprehensive study data available in
literatures, real-world applications and practical
experiments, we elucidate the formidable curse of
dimensionality and its adverse effects on model
generalization, computational resources and
interpretability. Furthermore, we delve into the arsenal
of dimensionality reduction techniques and feature
selection strategies, revealing the power of transforming
the data into actionable insights. This paper
demonstrates the tangible benefits of effectively
managing dimensions in machine learning, providing
practitioners with invaluable insights to harness the true
potential of their data. To validate the efficacy and
reliability of our proposed methodology, I conducted a
case study using a simple and informative dataset,
specifically focusing on Iris dataset.
Keywords :
Machine Learning, Informative and Simple Dataset, Dimensionality Reduction, PCA, LDA, Dashboards.