Linear Model Partitioning for Approximate Portfolio Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Electronic trading systems face challenges in efficiently managing and processing large volumes of data related to financial instrument positions, leading to increased computational resource demands and communication bandwidth limitations, which can result in slowed performance and higher costs.
Innovation Solution
A computer-implemented system that iteratively partitions large linear optimization models into smaller sub-models, allowing for independent solution of these sub-models using existing algorithms, thereby reducing the computational burden and enabling efficient data compression while maintaining key portfolio characteristics.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If large linear optimization models are processed using existing algorithms, then solution accuracy is maintained, but computational resource consumption increases and processing time extends
Solution Approach 1:
The patent divides a large linear optimization model into multiple smaller sub-models through iterative partitioning. Each sub-model can be solved independently using existing algorithms, reducing computational resource consumption and processing time while maintaining solution accuracy through aggregation of sub-model results
Solution Approach 2:
The patent transforms the problem from solving one large model in the original dimension to solving multiple smaller models in a partitioned dimension, then aggregating results. This dimensional transformation allows existing algorithms to work efficiently on smaller subsets while achieving the same overall solution quality
2Productivity
If computational resources are increased to handle large data volumes, then processing capability improves, but infrastructure costs increase
Solution Approach 1:
By segmenting the large optimization model into smaller sub-models, the patent enables processing with existing computational resources rather than requiring additional hardware or cloud infrastructure, thereby improving processing capability without increasing resource quantity
Solution Approach 2:
The patent creates multiple copies of the solving algorithm, each applied to a different sub-model. This allows parallel processing of smaller models using existing resources, effectively increasing processing capability without additional infrastructure investment
3Productivity
If data transmission volume is reduced through compression, then communication efficiency improves, but data integrity may be compromised
Solution Approach 1:
The patent extracts and removes redundant or less critical data elements during the model partitioning and compression process, reducing transmission volume while preserving the essential information needed to maintain data integrity in the compressed representation
Solution Approach 2:
The patent transforms the optimization model into a compressed format by changing parameters such as variable selection, constraint aggregation, and objective function simplification, achieving reduced data transmission volume while maintaining the reliability of the solution through controlled approximation
Data Source
AI summary
The disclosed embodiments related to multilateral portfolio compression using general large-scale linear optimization which pre-processes a model to decrease model size using domain knowledge to remove variables to reduce dimensionality, thereby making the model faster to solve and improving numerical characteristics. but it would not remove, for example, as much as half of the model, but rather a smaller fraction. The disclosed pre-processing enables an approximate solution for large, linear optimization models by automatically iteratively and selectively partitioning them into independently easily solvable sub-models. The sub-models are themselves linear optimization models, which can be solved with any preferred algorithm or library. The solutions for each sub-model are aggregated to obtain an acceptable, e.g., approximate, solution for a large model without solving the full model. At each iteration the disclosed embodiments will have a valid, feasible solution, if the user is satisfied before full convergence.