Random telegraph noise (RTN) shifts the threshold voltage (V_t) of 3-D NAND flash memory cells, making it a major cause of device malfunction. As device scaling continues, RTN has become an increasingly significant factor affecting device performance. The aim of this study is to develop a simulator that predicts the distribution of V_t shifts induced by RTN in scaled 3-D NAND flash memory. Previous RTN analysis methods rely heavily on numerous simulations or measurements, which are not only time-consuming but also limited in predicting the effects of device scaling on RTN-induced V_t shifts. To address these limitations, we developed a novel RTN Monte Carlo simulator that integrates a previously developed artificial neural network (ANN)-based machine learning (ML) model with a Markov process for trap occupancy states. Using this simulator, we comprehensively analyzed RTN effects in 3-D NAND devices with multiple traps, extracted the corresponding decay constants (λ), and modeled the dependence of λ on device physical parameters. The simulator provides flexibility in generating large-scale RTN data without the need for additional simulations or measurements, significantly reducing computation time while maintaining accuracy.