Classification of the hyperspectral images is a critical task for monitoring the environment, remote sensing and the agriculture. Selecting an appropriate classification algorithm that effectively handles the complex spectral and spatial information inherent in hyperspectral data is pivotal. This paper presents evaluation-based study of four prominent machine learning (ML) algorithms Support Vectors Machines, Multiple Layered Perceptron, Convolutional Neural Networks and the Random Forests (applied to the Indian Pines hyperspectral dataset). This study not only establishes a benchmark for algorithm performance on the Indian Pines dataset but also enhances understanding of algorithmic behaviour in hyperspectral image classification. The insights gained from this comparative analysis contribute to informed decision-making in algorithm selection and have implications for various applications requiring accurate and robust hyperspectral image classification.