Causal inference is important for discovering cause-and-effect relationships across disciplines, enabling insights beyond simple correlations. This paper reviews techniques like Graph Neural Networks (GNNs), Causal Forests, DEMATEL, (ISM) Interpretive Structural Modelling, (FCM) Fuzzy Cognitive Mapping, Bayesian Networks, Structural Equation Modelling (SEM), Directed Acyclic Graphs (DAGs), Granger Causality Analysis, and Markov Chains. These methods are compared across metrics such as conceptual framework, applicability, data requirements, model complexity, interpretability, accuracy, assumptions, and ease of use. Traditional techniques like SEM and Bayesian Networks excel in modeling latent variables, while DAGs clarify confounders. Machine learning-based methods like GNNs and Causal Forests handle high-dimensional data and heterogeneous effects. DEMATEL, ISM, and FCM are valuable for decision-making in systems with uncertainty. This review highlights the evolution from statistical to advanced computational approaches, emphasizing their growing importance in decision-making, policy evaluation, and complex system analysis across healthcare, economics, and AI.