Towards Leveraging AutoML for Sustainable Deep Learning: A Multi-Objective HPO Approach on Deep Shift Neural Networks
Researchers have developed a multi-objective hyperparameter optimization (HPO) approach for Deep Shift Neural Networks (DSNNs) to promote sustainable deep learning. This method combines multi-fidelity HPO with multi-objective optimization to balance model accuracy with energy consumption. Experiments show the approach can yield models with over 80% accuracy while significantly reducing computational costs, thereby facilitating more efficient and sustainable AI development. AI
IMPACT This research offers a method to develop more energy-efficient AI models, potentially reducing the environmental footprint of deep learning applications.