Building simulation (BS) increasingly relies on data-driven models that extract patterns directly from measured data. However, these models often conflate statistical dependency with causal relationship. The idea of a causal lens introduces structural causal diagrams and do-operators to distinguish true causations from spurious associations. The “causal lens” perspective highlights how confounding bias can arise in observational modeling and emphasizes the importance of extracting true causality from building data. This suggests that BS move beyond pattern replication to enable counterfactual reasoning, thereby supporting reliable decision-making.

Parameter estimation methods can be classified into (1) manual (trial-and-error), (2) numerical optimization (optimization, sampling), (3) Bayesian inference (Bayes filter, Bayesian calibration), and (4) machine learning (generative model). Bayesian calibration has been widely used because it can capture stochastic nature of uncertain parameters. However, the results of Bayesian calibration could be biased by (1) the prior distribution assumed by the expert's subjective judgment; (2) the likelihood function that cannot always describe the true likelihood; and (3) the posterior distribution approximation method, such as the Markov Chain Monte Carlo, which requires significant computation time. To overcome this, a new approach using a generator-regularized continuous conditional generative adversarial network (GRcGAN) is presented in this paper. Five target parameters of the DOE reference building model were selected. GRcGAN was trained to estimate uncertain parameters using simulated monthly electricity and gas use. GRcGAN can successfully estimate five uncertain parameters based on 1,000 training data points. The proposed approach presents a potential for stochastic parameter estimation.