Sparse Data Structure Compression with Mode Search Approximation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional delta compression methods fail to provide satisfactory results for sparse data structures, such as sparse matrices used in machine learning, due to large differences between neighboring numerical values, which defeat traditional compression schemes.
Innovation Solution
A data compression method utilizing mode search approximation, where the most repeated value (mode) is extracted and encoded with a significance map, complemented by a ternary tree processing algorithm for rapid mode approximation, and can operate in parallel with delta compression.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If conventional delta compression is applied to sparse data structures, then compression is performed, but compression efficiency deteriorates due to large differences between neighboring numerical values
Solution Approach 1:
The patent changes the compression parameter from delta encoding to mode-based encoding. Instead of storing differences between consecutive values, the system identifies the mode (most frequent value) in a window and stores deviations from this mode, which are typically smaller and more compressible for sparse data structures.
Solution Approach 2:
The patent introduces a dynamic windowing mechanism where the mode search operates over a sliding window of values. The window size and position are adjusted dynamically to capture local patterns in the data, allowing the compression to adapt to varying data characteristics throughout the sparse structure.
2Productivity
If mode search approximation is applied, then compression efficiency for sparse data structures is improved, but computational complexity increases
Solution Approach 1:
The patent divides the data into fixed-size windows and performs mode search independently on each window. This segmentation allows parallel processing of different windows and reduces the computational burden compared to searching the entire dataset, while still capturing local patterns effectively.
Solution Approach 2:
The patent uses approximation techniques where instead of finding the exact mode in every case, it employs heuristics and early termination strategies. When a sufficient number of matching values are found within a window, the search can be terminated early, providing good enough compression without exhaustive computation.
Data Source
AI summary
Embodiments are generally directed to compression for compression for sparse data structures utilizing mode search approximation. An embodiment of an apparatus includes one or more processors including a graphics processor to process data; and a memory for storage of data, including compressed data. The one or more processors are to provide for compression of a data structure, including identification of a mode in the data structure, the data structure including a plurality of values and the mode being a most repeated value in a data structure, wherein identification of the mode includes application of a mode approximation operation, and encoding of an output vector to include the identified mode, a significance map to indicate locations at which the mode is present in the data structure, and remaining uncompressed data from the data structure.