Key Challenges in Computational Methods
Luis G. Valerio · 2024
The many advantages of computational methods have gained increased interest across disciplines in the use of predictive analytic models to posit these tools with a crucial role in simulating and studying complex systems. However, to make advancements, the development and application of computational models are not without their challenges. Common and unique challenges will be discussed in this chapter. A central theme to key challenges encountered in model development relates to the complexity of models, a requisite of acquiring ample high-quality data, and the need for education and expertise to ensure proper use and interpretation of model outputs for scientists in the community. These challenges are discussed, including technical ones such as selection of features, scalability of models as the availability of data grows, and the wide range of model testing and validation methods to gain confidence and assess the skill and predictivity of the model. Practicalities are also covered such as generalizability and aging of models and translation of model output information for meaningful application.