MAP Estimation via Discrete DC Programming

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional MAP estimation techniques for graphical models, such as Stirling's approximation and continuous relaxation, result in low solution accuracy and poor interpretability, especially when dealing with small sample sizes and aggregated data.

Innovation Solution

The method transforms the MAP estimation problem into a minimum cost flow problem on a network and applies discrete DC programming to efficiently solve it without approximation, ensuring high accuracy and interpretability of the solution.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If conventional MAP estimation techniques (Stirling's approximation and continuous relaxation) are used, then the computation can be performed, but the solution accuracy and interpretability are low

Engineering Contradiction:

Improvesolution accuracyVSAvoidcomputation complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent replaces conventional approximation-based MAP estimation methods with discrete DC programming, substituting a mathematical optimization framework that guarantees exact solutions while maintaining computational efficiency through convex-concave decomposition

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent transforms the MAP estimation problem by changing the parameter representation to a flow-based formulation on a constructed graph, where the objective function is expressed as the difference of two convex functions (DC programming), enabling exact optimization without approximation

Inventive Principle:

Principle #35Parameter changes

2Loss of information

If conventional MAP estimation techniques are used, then the computation can be performed, but the solution interpretability is poor

Engineering Contradiction:

Improvesolution interpretabilityVSAvoidmethod complexity

Core Design Contradiction:

Loss of informationVSDevice complexity

Solution Approach 1:

The patent replaces continuous relaxation methods that produce fractional, hard-to-interpret values with discrete DC programming that naturally yields integer flow assignments, restoring interpretability through exact combinatorial optimization on a constructed graph

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Reliability

If discrete DC programming is used to solve MAP estimation, then solution accuracy and interpretability are improved, but the problem transformation complexity increases

Engineering Contradiction:

Improvesolution reliabilityVSAvoidproblem transformation complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent segments the MAP estimation problem into a structured graph construction phase followed by discrete DC programming optimization, where the graph divides variables and constraints into manageable nodes and edges, enabling systematic solution of the otherwise intractable problem

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS20230376564A1Estimation method, estimation apparatus and program

Publication Date: 2023.11.23 NT T INC

AI summary

An estimation method according to an embodiment is an estimation method that estimates a MAP solution of a CGM on a path graph, in which a computer executes: an input procedure that receives, as inputs, aggregate data and potentials of the CGM on the path graph; an estimation procedure that uses the aggregate data and the potentials to solve a MAP estimation problem of the CGM by a technique based on discrete DC programming and calculates a MAP estimation solution; and an output procedure that outputs the MAP estimation solution.