<![CDATA[Stars are celestial bodies that can be classified based on several characteristics, including temperature, luminosity, radius, magnitude, stellar color, and spectral class. Stars are generally grouped into six categories: brown dwarfs, red dwarfs, white dwarfs, main sequence stars, supergiants, and hypergiants. Stellar classification is important to astronomers because it can help identify new types of stars and improve our understanding of their composition, temperature, and evolutionary stages. This classification process can be done using the Nearest Centroid Neighbor (NCN) and k-Nearest Neighbor (k-NN) algorithms, by applying various distance measures such as Euclidean, Manhattan, Minkowski, Chebyshev, Cosine, Jaccard, and Hamming. This study aims to compare the performance between NCN and k-NN using these seven distance measures. The results show that Euclidean, Manhattan, and Minkowski distances produce a perfect performance of 100% in both algorithms. Chebyshev distance yielded perfect performance in k-NN but slightly lower in NCN with a performance of 92%. Thus, the k-NN algorithm provides superior performance compared to the NCN algorithm in the stellar classification process.]]>
Copyrights © 2026
