Performance analysis is utmost important with the purpose of propelling the development of both software and hardware systems. By conducting a comprehensive study, it is possible to identify bottlenecks in the system and architecture, obtain essential information for selecting frameworks and platforms, and ultimately result in improvements in performance. Performance analysis is major step towards performance optimizations. Various methods of optimization exist in order to achieve better performance of various machine learning algorithms. In this paper, performance of proposed forecasting models is compared with various existing models. The performance of proposed models Improved Whale Optimization Based Forecasting Model (IWOFM) and Adaptive Gradient Based Forecasting Model (AGBFM) is compared after the final iteration of the forecasting models with LSSVM-PSO, LSSVM-ACO and LSSVM-WOA using the similar dataset scenarios.