Abstract of the talk// Multi-objective optimization techniques play a crucial role in addressing complex problems where multiple, often conflicting, objectives need to be optimized simultaneously. In the context of the hyperspectral dataspace, such techniques are particularly valuable due to the high-dimensional nature of the data and the need to balance various performance metrics like classification accuracy, computational efficiency, and feature selection. 

Tree-structured Parzen Estimator (TPE) is a Bayesian optimization method widely used for hyperparameter tuning in machine learning. Traditionally applied to single-objective problems, recent advancements have extended TPE to multi-objective optimization, making it suitable for scenarios where multiple criteria must be considered. For hyperspectral dataspace, TPE can be employed to optimize the hyperparameters of machine learning models used for classification or regression, balancing objectives such as accuracy, precision, recall, and computational efficiency.

NSGA-III, on the other hand, is a state-of-the-art evolutionary algorithm designed specifically for multi-objective optimization. It excels in maintaining a diverse set of non-dominated solutions, enabling the exploration of trade-offs between conflicting objectives. NSGA-III has been successfully applied to various domains and enables the exploration of the trade-off space between different objectives, providing decision-makers with a range of optimal solutions. In the hyperspectral dataspace, NSGA-III has been applied to tasks such as band selection, feature extraction, and sparse decomposition. For example, it has been used to optimize the selection of spectral bands, maximizing information content while minimizing redundancy, thereby enhancing the efficiency of subsequent analyses. 

In his talk, Parth will explore the principles of TPE and NSGA-III, their adaptability to challenges in hyperspectral dataspace, and their potential to enhance performance in real-world applications, such as remote sensing and environmental monitoring.