Toward Sustainable AI: Principles, Optimization Strategies and a Lightweight MLP Application in Air Quality Forecasting

Key Points

• Air quality forecasting accuracy significantly improves with sustainable AI optimization strategies.
• The use of a lightweight multilayer perceptron (MLP) model consistently outperforms conventional algorithms in predictive accuracy.
• Observational analysis on air quality data illustrates the effectiveness of sustainability-focused AI principles for accurate forecasting.
• Implementing these strategies may enhance environmental decision-making, highlighting the need for broader application in various sectors.

Abstract