This study compares and contrasts two cutting-edge methods for determining the authorship of Urdu text: Hyper-Tuned CNN-based deep learning and Hybrid Ensemble Learning-based method. The latter method uses ensemble SVM with boosted algorithms, such as Gradient Boosting (GBC), Catboosting (CBC), and XGBoosting (XGB) classification models. The purpose of the study is to assess how well these methods work at locating the Urdu-language author of a given text document. In comparison to the Hybrid Ensemble Learning technique using boosted SVM algorithms, the experimental results demonstrate that the HyperTuned CNN based deep learning strategy provides higher outcomes in terms of accuracy, precision, and recall. These results indicate that the Hyper-Tuned CNN based deep learning methodology is an effective method for determining who wrote a piece of Urdu text. It also suggests that this method may be useful for other text categorization problems. The study also emphasises the significance of comparison studies in assessing the efficiency of various machine learning approaches for text classification tasks. It is necessary to conduct additional study to examine the applicability of these strategies in additional languages and to determine whether they can be used to various text classification tasks.