This study presents a novel method for predicting the performance and emissions of a Homogeneous Charge Compression Ignition-Direct Injection (HCCI-DI) engine fuelled by waste cooking oil biodiesel (WCOB) blended with Aluminum oxide (Al2O3) and Ferric chloride (FeCl3) nano additives and premixed with gasoline. The test fuels considered were pure diesel, WCOB, B50 (a 50–50 blend of diesel and WCOB), and the nano-additives at concentrations of 50 ppm and 100 ppm. All experiments were performed on a 4.4 kW HCCI-DI engine operating at 1500 rpm. The results revealed that, utilization of WCOB showed a substantial reduction in emissions. Hydrocarbon (HC), carbon monoxide (CO) and smoke emissions decreased by 54.17 %, 50 %, and 22.69 %. Nevertheless, oxides of nitrogen (NOx) emissions increased by 18.32 %. When introducing gasoline (20 %) as a premix in HCCI-DI engines, a favourable shift in emission and efficiency metrics was observed. The brake thermal efficiency (BTE) increased 4.23 %. Moreover, NOx and smoke emissions decreased by 4.3 % and 39.21 %. Furthermore, compared to conventional diesel-fueled combustion, the integration of the nano additives manifested promising results. After introducing Al2O3 into neat fuel, The BTE improved by 11.27 % for direct injection (DI) combustion and 18.31 % for HCCI-DI combustion. Additionally, there was a marked decrease in exhaust emissions. FeCl3 nano additive into the test fuel significantly reduced HC, CO, and smoke emissions. Furthermore, the random forest machine learning approach demonstrated an extensive accuracy in forecasting both engine performance and emissions for the HCCI-DI engine. The innovative combination of cutting-edge machine learning techniques and combustion technology used. This model to understand the complicated correlations between input parameters and engine outputs and account for the system's intrinsic non-linearities and interactions. This dramatically improves standard prediction approaches, often assuming linear correlations or disregarding variable interdependencies.