Coordinate Matrix Machine: A Human-level Concept Learning to Classify Very Similar Documents

TL;DR

The paper tackles the problem of classifying highly similar structured documents with minimal labeled data. It introduces the Coordinate Matrix Machine ($CM^2$), a one-shot, coordinate-based approach that constructs a keyword-position matrix and uses a distance-based matcher (Manhattan distance) to classify documents, optimized for CPU-only inference and interpretability. The work demonstrates that $CM^2$ surpasses traditional vectorizers and DL models in accuracy, speed, and robustness on bank statements while aligning with Green AI principles. This method offers a practical, auditable solution for high-stakes domains such as finance and law that require transparency and low environmental impact.

Abstract

Human-level concept learning argues that humans typically learn new concepts from a single example, whereas machine learning algorithms typically require hundreds of samples to learn a single concept. Our brain subconsciously identifies important features and learns more effectively. \vspace*{6pt} Contribution: In this paper, we present the Coordinate Matrix Machine (CM$^2$). This purpose-built small model augments human intelligence by learning document structures and using this information to classify documents. While modern "Red AI" trends rely on massive pre-training and energy-intensive GPU infrastructure, CM$^2$ is designed as a Green AI solution. It achieves human-level concept learning by identifying only the structural "important features" a human would consider, allowing it to classify very similar documents using only one sample per class. Advantage: Our algorithm outperforms traditional vectorizers and complex deep learning models that require larger datasets and significant compute. By focusing on structural coordinates rather than exhaustive semantic vectors, CM$^2$ offers: 1. High accuracy with minimal data (one-shot learning) 2. Geometric and structural intelligence 3. Green AI and environmental sustainability 4. Optimized for CPU-only environments 5. Inherent explainability (glass-box model) 6. Faster computation and low latency 7. Robustness against unbalanced classes 8. Economic viability 9. Generic, expandable, and extendable

Figures (3)

• Figure 1: The Distribution of Statements for Each Bank
• Figure 2: Performance of Different Vectorizer with Every Classifier Compared Against CM$^2$ for Varying Training Data Set
• Figure 3: Effect of Different Maximum Penalty on CM$^2$ Performance

Theorems & Definitions (6)

• Example 1.1
• Example 1.2
• Example 3.1
• Example 3.2
• Definition 3.1
• Example 3.3