Trustworthy Tree-based Machine Learning by $MoS_2$ Flash-based Analog CAM with Inherent Soft Boundaries

TL;DR

This paper introduces a novel hardware-software co-design using $MoS_2$ Flash-based analog CAM with soft boundaries, enabling robust, efficient, and explainable tree-based AI models resistant to device variations and adversarial attacks.

Contribution

It presents a new hardware-software co-design approach with $MoS_2$ analog CAM that inherently supports soft decision boundaries, improving robustness and accuracy of tree-based models.

Findings

Achieved 96% accuracy on WDBC dataset with analog CAM.

Model shows only 0.6% accuracy drop under 10% device variation on MNIST.

Outperforms traditional decision trees in robustness against device variation.

Abstract

The rapid advancement of artificial intelligence has raised concerns regarding its trustworthiness, especially in terms of interpretability and robustness. Tree-based models like Random Forest and XGBoost excel in interpretability and accuracy for tabular data, but scaling them remains computationally expensive due to poor data locality and high data dependence. Previous efforts to accelerate these models with analog content addressable memory (CAM) have struggled, due to the fact that the difficult-to-implement sharp decision boundaries are highly susceptible to device variations, which leads to poor hardware performance and vulnerability to adversarial attacks. This work presents a novel hardware-software co-design approach using $MoS_2$ Flash-based analog CAM with inherent soft boundaries, enabling efficient inference with soft tree-based models. Our soft tree model inference experiments on $MoS_2$ analog CAM arrays show this method achieves exceptional robustness against device variation and adversarial attacks while achieving state-of-the-art accuracy. Specifically, our fabricated analog CAM arrays achieve $96\%$ accuracy on Wisconsin Diagnostic Breast Cancer (WDBC) database, while maintaining decision explainability. Our experimentally calibrated model validated only a $0.6\%$ accuracy drop on the MNIST dataset under $10\%$ device threshold variation, compared to a $45.3\%$ drop for traditional decision trees. This work paves the way for specialized hardware that enhances AI's trustworthiness and efficiency.