In recent years, in the context of tabular dataTabular data classificationClassification, the emergence of the transformer architecture has led to deep learning methods yielding better results than conventional tree-based models. Most of these findings attribute the success of deep learning to the expressive feature constructionFeature construction capabilities of neural networks. Nonetheless, in real-world practice, manually designed high-order features using traditional machine learning methods are still widely used because features based on neural networks can be prone to overfitting. In this paper, we propose a genetic programming-based feature engineeringFeature engineering algorithm to automate the feature constructionFeature construction process through trial and improvement. Importantly, genetic programming provides an opportunity to optimize symbolic models, which gradient-based methods often find hard to optimize. On a large-scale classificationClassification benchmark involving 130 datasets, the experimental results demonstrate that the proposed method outperforms existing, fine-tuned state-of-the-art tree-based and deep-learning-based classificationClassification algorithms.