Hierarchical Posterior Distribution Calculator for Fast Parameter Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for estimating parameter values in simulations with a large number of parameters require excessive time, leading to impractical computation times as the number of parameters increases exponentially, potentially preventing timely estimation.
Innovation Solution
A data calculation device and method utilizing hierarchical models, where a high-level model simulates a subject and low-level models simulate parts of the subject, allowing for the estimation of parameter values in a hierarchical structure, reducing the number of parameters in each model and facilitating faster computation by using high-level model outputs as inputs for low-level model simulations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If parameter values are estimated using existing methods for simulations with a large number of parameters, then the estimation accuracy can be maintained, but the computation time increases exponentially
Solution Approach 1:
The patent divides the simulation model into hierarchical levels (high-level model and low-level models). The high-level model estimates parameters for groups of subsystems, while low-level models estimate parameters for individual subsystems. This segmentation reduces the number of parameters estimated simultaneously, thereby reducing computation time while maintaining estimation accuracy through iterative refinement across levels.
2Adaptability or versatility
If the number of parameters for estimation is increased to cover a large-scale simulation subject, then the coverage and applicability improve, but the computation time becomes impractically long
Solution Approach 1:
The simulation subject is divided into multiple subsystems organized in a hierarchical structure. The high-level model handles parameters for multiple subsystems simultaneously, while low-level models handle individual subsystem parameters. This allows the system to maintain broad simulation coverage across large-scale subjects while keeping the computational burden manageable through the hierarchical division of parameter estimation tasks.
Solution Approach 2:
The patent introduces a hierarchical dimension to the parameter estimation process, organizing parameters into multiple levels (high-level and low-level). This dimensional organization allows the system to handle a large total number of parameters by distributing them across hierarchical levels, thereby maintaining simulation coverage while reducing the computational complexity at each estimation step.
3Device complexity
If a single model is used to simulate the entire subject, then the model structure remains simple, but the number of parameters increases exponentially leading to long computation times
Solution Approach 1:
Instead of using a single model for the entire subject, the patent segments the model into a high-level model and multiple low-level models. The high-level model has a simpler parameter set that covers multiple subsystems, while low-level models have simpler individual parameter sets. This segmentation maintains overall model structure simplicity while reducing the exponential growth of parameters by distributing them across hierarchical levels.
Data Source
AI summary
A data calculation device including a high-level posterior distribution data calculator for calculating data indicating a posterior distribution of parameter values of a high-level model that simulates a subject, based on target data for outputs from the high-level model; and a low-level posterior distribution data calculator for calculating, for a low-level model that simulates a portion of the subject and that outputs parameter values of the high-level model, data indicating a posterior distribution of parameter values of the low-level model, based on the data indicating a posterior distribution of parameter values of the high-level model.