Energy Efficient Medical Data Dimensionality Reduction using Optimized Principal Component Analysis

INTRODUCTION: The method of minimizing the number of random variables or attributes from the enormous data set is the reduction of dimensionality. The space available for storing the database is therefore minimized by decreasing the scale of the features. OBJECTIVES: The PCA algorithm is used...

Descripción completa

Guardado en:
Detalles Bibliográficos
Autores principales: S. Sophia, K. Thanammal, S. Sujatha
Formato: article
Lenguaje:EN
Publicado: European Alliance for Innovation (EAI) 2022
Materias:
Q
Acceso en línea:https://doaj.org/article/f7e660bb57134f518b3750de86fc828b
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
Descripción
Sumario:INTRODUCTION: The method of minimizing the number of random variables or attributes from the enormous data set is the reduction of dimensionality. The space available for storing the database is therefore minimized by decreasing the scale of the features. OBJECTIVES: The PCA algorithm is used to achieve dimensional reduction by deep learning to recover image characteristics. This approach is designed to reduce the dimensionality of such datasets, improve interpretability while minimizing the loss of information.METHODS: The dimensionality reduction of the method by using optimized PCA algorithm. The input data set can be reducing the dimension by using PCA algorithm. The tree seed optimization algorithm (TSO) can be utilized to select the optimal data’s in PCA algorithms. After completing the TSO-PCA the new data set are created by the reduced dimensions. RESULTS: The input data and images are used to reduce the dimension based on the TSO-PCA algorithms. The simulations for obtaining the results were carried out using python. The results of the feature dimensionality reduction on DIABETES dataset and Indian pines dataset. CONCLUSION: The best data for the data collection, the TSO algorithm is used and the PCA algorithm is used to minimize the dimensions. The suggested method is better than the existing method compared to the linear, kernel, random basic function, and polynomial for evaluating the outcome and discussion. In order to improve accuracy in future work, we will continue research and try to find more advanced techniques for this problem.