Associative Matrix Compression via Bit Plane Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Associative memory systems face challenges in scaling due to geometric growth with the number of inputs, limiting their application in complex real-world problems, and there is a need for lossless compression of large association matrices while allowing random access.
Innovation Solution
The method involves compressing association matrices by selectively representing segments with non-zero counts as bit planes, generating maps for non-zero data, and reorganizing segments into continuous ranges for efficient storage and retrieval.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If associative memories are used to solve real world problems, then pattern matching and identification capabilities are improved, but geometric scaling limitations prevent support for applications at the required scale of complexity
Solution Approach 1:
The association matrix is divided into multiple segments, where each segment stores a portion of the association data. This segmentation allows the system to handle large-scale applications by distributing the computational load across multiple manageable segments, thereby overcoming the geometric scaling limitations of traditional associative memories while maintaining pattern matching capabilities.
2Quantity of substance
If the association matrix is compressed to reduce memory usage, then storage efficiency is improved, but random access capability may be degraded
Solution Approach 1:
The patent introduces a segmented organizational structure that adds a hierarchical dimension to the association matrix storage. By organizing data into segments with associated metadata structures, the system achieves lossless compression while maintaining random access capability through direct segment addressing and efficient segment-level navigation.
3Quantity of substance
If segments are reorganized into continuous ranges for efficient storage, then storage density is improved, but access patterns may become more complex
Solution Approach 1:
The patent pre-organizes segments into continuous ranges during the initialization and data loading phase. By establishing contiguous segment arrangements upfront and maintaining segment metadata structures, the system achieves high storage density while simplifying access patterns during operation, as the continuous ranges enable efficient sequential and random access without complex reorganization during runtime.
Data Source
AI summary
Associative matrix compression methods, systems, computer program products and data structures compress an association matrix that contains counts that indicate associations among pairs of attributes. Selective bit plane representations of those selected segments of the association matrix that have at least one count is performed, to allow compression. More specifically, a set of segments is generated, a respective one of which defines a subset, greater than one, of the pairs of attributes. Selective identifications of those segments that have at least one count are stored. The at least one count that is associated with a respective identified segment is also stored as at least one bit plane representation. The at least one bit plane representation identifies a value of the at least one associated count for a bit position of the count that corresponds to the associated bit plane.