Common phenomena associated with alluvial river is its meandering behavior A meander forms when the flowing water in a stream alters its speed, leading to erosion of sediment from the outer edge of a bend and subsequent deposition on the inner edge of the bend. Sinuosity is a major parameter use to classify a river is meandering or not. Sinuosity is defined as the ratio of the stream length to valley. A meandering compound channel consists of a primary channel responsible for transporting base flow and common runoff up to its maximum capacity, alongside a floodplain on one or both sides, which accommodates excess water during flood events. Numerous studies have been conducted to analyze the hydraulic behavior of flow in sinuous compound channels, leading to the development of various methodologies. In general, this method either have numerical solution of differential equation or need long computation. This study focuses on predicting flow in meandering compound channels through the application of Machine Learning (ML) techniques, specifically Artificial Neural Network Particle Swarm Optimization (ANN-PSO) and M5 Tree models. These models consider various non-dimensional parameters, including depth ratio, channel sinuosity, bed slope, width ratio, and discharge ratio. The findings indicate that both the ANN- PSO and M5 Tree models are capable of accurately predicting discharge (QP), achieving a coefficient of determination (R2) value above 0.75 and a mean absolute percentage error (MAPE) below 10% for both training and testing datasets. However, the M5 Tree model demonstrates superior predictive accuracy compared to the ANN PSO model.