Bitmap Decision Tree Classifier Construction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing decision tree algorithms are inefficient in handling large datasets due to the high computational cost of constructing count tables for predictor-target pairs, which limits their scalability.
Innovation Solution
A computer-implemented method that generates bitmaps for each tree node by intersecting parent node bitmaps with predictor and target bitmaps, counting bits to efficiently determine predictor-target pairs, and recursively building decision trees within a database system, leveraging memory management and parallel execution.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional decision tree algorithms construct count tables for predictor-target pairs, then classification accuracy can be achieved, but computational resources and time requirements increase significantly
Solution Approach 1:
The patent replaces the traditional mechanical counting process with a bitmap-based parallel processing system. Instead of sequentially counting predictor-target pairs, the system uses bitmaps to represent data and performs parallel intersection operations to count pairs, significantly reducing computational resource requirements while maintaining accuracy
Solution Approach 2:
The patent changes the data representation parameter from traditional tabular format to bitmap format. This parameter change enables efficient parallel processing and reduces the computational complexity of constructing count tables, allowing the system to handle large datasets with limited computational resources
2Quantity of substance
If traditional decision tree algorithms process large datasets, then comprehensive analysis is achieved, but processing time increases significantly
Solution Approach 1:
The patent segments the counting process into parallel operations. By dividing the data into bitmap representations and using parallel intersection operations across multiple processors, the system can process large datasets simultaneously rather than sequentially, dramatically reducing processing time
Solution Approach 2:
The patent substitutes sequential mechanical counting with parallel bitmap operations. This substitution enables the system to handle large datasets efficiently by performing multiple counting operations simultaneously across different predictor-target pairs, reducing overall processing time
3Measurement precision
If decision tree algorithms scale to large datasets, then model accuracy improves, but computational complexity increases
Solution Approach 1:
The patent changes the fundamental parameter of data representation to bitmaps, which enables efficient scaling to large datasets. This parameter change reduces computational complexity by allowing parallel processing operations that can handle large data volumes without proportionally increasing computational requirements
Solution Approach 2:
The patent creates a universal bitmap-based system that can handle various dataset sizes and complexities through the same underlying mechanism. The bitmap intersection and counting operations serve multiple functions including data filtering, counting, and pattern recognition, reducing overall system complexity while maintaining accuracy across different data scales
Data Source
AI summary
A method, system, and computer program product for counting predictor-target pairs for a decision tree model provides the capability to generate count tables that is quicker and more efficient than previous techniques. A method of counting predictor-target pairs for a decision tree model, the decision tree model based on data stored in a database, the data comprising a plurality of rows of data, at least one predictor and at least one target, comprises generating a bitmap for each split node of data stored in a database system by intersecting a parent node bitmap and a bitmap of a predictor that satisfies a condition of the node, intersecting each split node bitmap with each predictor bitmap and with each target bitmap to form intersected bitmaps, and counting bits of each intersected bitmap to generate a count of predictor-target pairs.