Although effectively managing construction costs is crucial for large-scale and high-rise construction projects, achieving efficiency and accuracy can be challenging. To improve the accuracy of the construction cost prediction model, this study presents an optimal combination of influential factors based on a correlation analysis of construction costs by type. The study establishes five models, including the top three models and the total area obtained through correlation analysis for each construction type, and utilizes the artificial neural network method to predict construction costs and compare the predictive performance of each model. Analysis shows that the combination of factors with vertical characteristics of the building yields a lower average error rate than other influencing factors. This finding is expected to help improve the combination of influencing factors for construction cost prediction models in future studies.