Abstract: The rapid and precise identification of the pulse-like ground motions is a key challenge that perplexes both the academic and engineering communities. The quantitative identification methods can overcome the empirical limitations of manual identification. However, the traditional quantitative identification methods suffer from inconsistencies in the identified results, limited applicability, and difficulties in simultaneously determining the accurate pulse periods. In response, a problem-targeted fusion learning rule is established, combined with a convolutional neural network (CNN) model, to develop a novel method to synchronously identify pulse-like ground motions and their pulse periods. This learning rule integrates multiple traditional typical identification methods based on different identification principles, thereby eliminating the cumbersome manual labeling process. It employs 30000 ground motion data from arbitrary directions worldwide for training and validation, resulting in three problem-targeted CNN models named the Strict, General, and T_P identification models. To address the issue of insufficient temporal input information for ground motions leading to weak model generalization capability, the input structure of the CNN model is optimized, and the ST-CNN model is proposed, incorporating the S-transform layer to convert ground motion time series to time frequency, thereby enhancing frequency domain distribution information and further improving the identification accuracy. The results indicate that the Strict model can strictly differentiate between the pulse-like and non-pulse-like ground motions, with the results consistent with those of other methods. The General model can identify more pulse-like ground motions and has broader applicability. The T_P model accurately identifies pulse periods and can be used in conjunction with the aforementioned models to synchronously output the identified results. The proposed problem-targeted fusion learning rule can also be extended to other engineering fields and other machine learning models, and the established identification method can provide scientific guidance for the study on the pulse-like ground motions.