A new study published on arXiv explores the prediction of energy consumption for Java methods. Researchers found that static code metrics alone are poor predictors of energy usage, yielding R2 values close to zero. However, incorporating method execution time as a dynamic input significantly improved accuracy, achieving R2 values up to 0.46. The study identified execution time, internal method calls, and cyclomatic complexity as the most influential factors in predicting energy consumption. AI
IMPACT This research could lead to more energy-efficient software development practices by enabling early detection of energy-intensive code.
RANK_REASON The cluster contains an academic paper detailing a research study.
AI-generated summary · Google Gemini · from 2 sources. How we write summaries →
