Simple Training Algorithms
Category: AI
Select the simplest and most energy-efficient training algorithm that meets the requirements of the task. In many cases, traditional or symbolic approaches (e.g., decision trees, linear models, or rule-based systems) can offer sufficient performance without the high training cost and energy consumption associated with complex models such as deep learning architectures. This pattern promotes thoughtful algorithm selection by encouraging developers and data scientists to assess whether simpler models or hybrid techniques are more appropriate for a given problem, especially when resource constraints or sustainability goals are part of the system requirements. Avoid defaulting to computationally intensive models when lighter alternatives offer comparable results with lower carbon impact.