Magnetic skyrmions, characterized by their topological properties, serve as core components for developing next-generation non-volatile memory devices that demand high density, high speed, and low power consumption. Their formation arises from the Dzyaloshinskii-Moriya interaction (DMI), enabled by non-centrosymmetric structures. Two-dimensional Janus magnetic materials, which inherently break spatial inversion symmetry, readily generate strong DMI, providing an ideal platform for skyrmion generation and novel racetrack memory applications. Within this field, identifying systems with a high Curie temperature ($T_{\rm{C}}$) is crucial, as it directly governs magnetic property stability and application potential under high-temperature conditions. This study integrated literature and open-source databases to construct a dataset of 16, 880 ABC-type two-dimensional materials. Utilizing stoichiometric ratios, intrinsic elemental properties, and electronic structure features as descriptors, four machine learning models—Random Forest (RF), Gradient Boosting Decision Tree (GBDT), Extreme Gradient Boosting (XGBoost), and Extra Trees (ET)—were employed for $T_{\rm{C}}$ prediction. Model performance was evaluated via ten-fold cross-validation, revealing that the XGBoost model exhibited superior prediction accuracy and generalization capability. Leveraging this model, $T_{\rm{C}}$ was predicted for 4, 024 unexplored two-dimensional Janus materials. This screening identified 54 promising candidates possessing thermal stability, high magnetic moment, and a $T_{\rm{C}}$ exceeding 300 K. To verify reliability, four candidate systems (EuFeO, GdKTi, DyFeTb, ErFeGd) were randomly selected for theoretical validation using first-principles calculations combined with the Heisenberg model. For systems exhibiting strong correlation effects (containing d-orbital electrons), the Hubbard U parameter was incorporated to describe on-site Coulomb repulsion. Exchange coupling constants were derived using the VASP software package. Subsequently, $T_{\rm{C}}$ values were calculated via classical Monte Carlo simulations performed using the MCSOLVER program. Results demonstrate that the mean absolute error (MAE) of the predicted $T_{\rm{C}}$ agrees well with the model calculations for EuFeO and GdKTi, while larger deviations were observed for DyFeTb and ErFeGd. Nevertheless, the calculated $T_{\rm{C}}$ values for all four candidates surpass room temperature. This work establishes a new computational framework for the efficient screening of high-performance two-dimensional Janus magnetic materials, contributing to the advancement of magnetic storage technologies.