Abstract: Accurate profit forecasting is critical for small-scale pharmacies, particularly in resource-constrained environments where financial decisions must be both timely and data-informed. This study investigates the predictive performance of sixteen regression models for daily profit forecasting using transactional data collected from a single local pharmacy in Sabha, Libya, over a 14-month period. An exploratory data analysis revealed strong right-skewed distributions in sales, cost, and profit, as well as pronounced temporal patterns, including seasonal peaks during spring and early summer and weekly profit clustering around weekends. After outlier treatment using the interquartile range method. A total of sixteen regression models were developed and evaluated, encompassing linear models (Linear, Ridge, Lasso, ElasticNet), tree-based models (Decision Tree, Random Forest, Extra Trees, Gradient Boosting, AdaBoost), proximity-based models (K-Nearest Neighbors), kernel-based models (Support Vector Regression), and neural architectures (Multi-Layer Perceptron, Convolutional Neural Network, Long Short-Term Memory, Gated Recurrent Unit). The models were assessed using Mean Absolute Error, Mean Squared Error, Root Mean Squared Error, and the R-squared score. The results consistently showed that tree-based ensemble models—particularly Extra Trees and LightGBM—achieved the highest accuracy, with R² values of 0.978 and 0.975 respectively, significantly outperforming neural and linear models. Learning curves and residual plots further confirmed the superior generalization and robustness of these models. We acknowledge that the dataset size (424 records) and the deterministic relationship between sales, costs, and profit influence these metrics. The study highlights the importance of model selection tailored to domain-specific data characteristics and suggests that well-tuned ensemble methods may offer reliable, interpretable, and scalable solutions for profit forecasting in simialr low-resource retail environments. However, broad claims of usefulness for all low-resource settings should be tempered by the limited scope of this dataset. Future work should consider longer-term data and external economic indicators to further improve model reliability, and focus on operational deployment strategies, investigating how these models can be integrated into daily pharmacy workflows despite real-time data constraints.

This study investigates logistic regression, linear support vector machine, multinomial Naive Bayes, and Bernoulli Naive Bayes for classifying Libyan dialect utterances gathered from Twitter. The dataset used is the QADI corpus, which consists of 540,000 sentences across 18 Arabic dialects. Preprocessing challenges include handling inconsistent orthographic variations and non-standard spellings typical of the Libyan dialect. The chi-square analysis revealed that certain features, such as email mentions and emotion indicators, were not significantly associated with dialect classification and were thus excluded from further analysis. Two main experiments were conducted: (1) evaluating the significance of meta-features extracted from the corpus using the chi-square test and (2) assessing classifier performance using different word and character n-gram representations. The classification experiments showed that Multinomial Naive Bayes (MNB) achieved the highest accuracy of 85.89% and an F1-score of 0.85741 when using a (1,2) word n-gram and (1,5) character n-gram representation. In contrast, Logistic Regression and Linear SVM exhibited slightly lower performance, with maximum accuracies of 84.41% and 84.73%, respectively. Additional evaluation metrics, including log loss, Cohen’s kappa, and Matthew’s correlation coefficient, further supported the effectiveness of MNB in this task. The results indicate that carefully selected n-gram representations and classification models play a crucial role in improving the accuracy of Libyan dialect identification. This study provides empirical benchmarks and insights for future research in Arabic dialect NLP applications.