This paper presents a novel time series forecasting method based on fuzzy cognitive maps (FCMs) induced by intuitionistic fuzzy set integrated with fuzzy C-means clustering and Particle Swarm Optimization (PSO). As FCM has inherent characteristics of scalability and adaptability, it is used in developed model and includes non-determinacy by using induced fuzzy set that are constructed from intuitionistic fuzzy sets (IFSs). The proposed approach aims to increase the precision in forecasting outputs by leveraging the strengths of each individual component in the presence of uncertainty and non-determinacy. FCMs are utilized to capture the underlying causal relationships in time series data, providing a dynamic representation of system behaviors. Fuzzy C-means clustering is used to create IFSs, enabling more effective handling of uncertainty and imprecision in the data. PSO in the proposed forecasting model is used to fine-tune the parameters of FCM, optimizing the learning process and improving forecasting performance. The proposed hybrid model of time series forecasting is evaluated on time series datasets of market price of State Bank of India and enrolments of the University of Alabama to demonstrate its superior execution compared to existing methods of fuzzy time series forecasting.