Forecasting of palm oil imports to India has gained significant prominence in contemporary times due to huge exchequer is spending on vegetable oil imports. The government is keen on reducing imports. The quantity to be imported in future years is utmost important to make any policies or programs to enhance the oilseeds production. For forecasting ARIMA models have been the most widely used technique during the last few decades. When the assumption of homoscedastic error variance is violated then ARCH/GARCH models are applied to capture the changes in the conditional variance of the time-series data. The machine learning techniques, i.e., ANN and SVR, can also be applied in the field of forecasting of real time-series data successfully as an alternative to the traditional forecasting models as these are data-driven models and could capture nonlinearities existing in the data. The present study analyzed monthly time series data of palm oil import volume (thousand tonnes) from the world to India from April 2007 to March 2023. It is clear from the results that the machine learning models, viz, SVR and ANN, outperformed the traditional time series models (GARCH and ARIMA) with the least RMSE, MAPE, Theil's U statistic and the highest CDC values for both training and testing datasets. Empirical results revealed that SVR is the best model for forecasting palm oil import volume compared to all other models.