ETHEREAL: Energy-efficient and High-throughput Inference using Compressed Tsetlin Machine

2025-02-08

AI Summary

To address model redundancy in Tsetlin Machines (TMs) that degrades energy efficiency and inference throughput in TinyML scenarios, this paper proposes a structural sparsification-based TM compression training method. Our approach introduces two key innovations: (1) leveraging the literal cancellation property between positive and negative clauses to achieve logic-pattern-level compression; and (2) incorporating an exclusion mechanism for non-critical automaton states, enabling fine-grained structural sparsification under discriminative capability constraints. Evaluated on the STM32F746G-DISCO platform, the method reduces model size by 87.54%, accelerates inference over Binary Neural Networks (BNNs) by more than 10×, and significantly lowers energy consumption. Memory footprint remains substantially smaller than that of Random Forests, while classification accuracy degradation is negligible. This work establishes a novel paradigm for efficient TM deployment on resource-constrained edge devices.

Technology Category

• Machine Learning: Learning on the Edge & Model Compression
• Data Mining & Knowledge Management: Data Compression
• Constraint Satisfaction and Optimization: Satisfiability Modulo Theories

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Security and Privacy: Security and privacy of machine learning and AI applications
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

The Tsetlin Machine (TM) is a novel alternative to deep neural networks (DNNs). Unlike DNNs, which rely on multi-path arithmetic operations, a TM learns propositional logic patterns from data literals using Tsetlin automata. This fundamental shift from arithmetic to logic underpinning makes TM suitable for empowering new applications with low-cost implementations. In TM, literals are often included by both positive and negative clauses within the same class, canceling out their impact on individual class definitions. This property can be exploited to develop compressed TM models, enabling energy-efficient and high-throughput inferences for machine learning (ML) applications. We introduce a training approach that incorporates excluded automata states to sparsify TM logic patterns in both positive and negative clauses. This exclusion is iterative, ensuring that highly class-correlated (and therefore significant) literals are retained in the compressed inference model, ETHEREAL, to maintain strong classification accuracy. Compared to standard TMs, ETHEREAL TM models can reduce model size by up to 87.54%, with only a minor accuracy compromise. We validate the impact of this compression on eight real-world Tiny machine learning (TinyML) datasets against standard TM, equivalent Random Forest (RF) and Binarized Neural Network (BNN) on the STM32F746G-DISCO platform. Our results show that ETHEREAL TM models achieve over an order of magnitude reduction in inference time (resulting in higher throughput) and energy consumption compared to BNNs, while maintaining a significantly smaller memory footprint compared to RFs.

Problem

• Develops compressed Tsetlin Machine models.
• Enables energy-efficient, high-throughput ML inferences.
• Reduces model size with minimal accuracy loss.

Innovation

• Compressed Tsetlin Machine models
• Energy-efficient inference approach
• High-throughput TinyML applications