Neural Network Input Ordering for Classification Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional machine learning techniques fail to accurately classify data due to the arbitrary ordering of input values, which affects the accuracy of neural network outputs, leading to poor classification results.
Innovation Solution
A method that determines an optimal input order for numerical values in a neural network using a reference pattern, updates the reference values based on input errors, and rearranges data to maximize similarity with the reference pattern, thereby improving classification accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If input values are arranged in arbitrary order, then the neural network processing is simple, but the classification accuracy deteriorates
Solution Approach 1:
The patent applies preliminary action by determining the optimal input order before the neural network processing begins. The determining unit establishes the sequence in which input values should be arranged based on their contribution to classification accuracy, so that when the neural network receives the data, it is already optimally ordered, preventing accuracy deterioration without adding complexity during the actual processing phase
Solution Approach 2:
The patent changes the parameter of input value ordering from arbitrary to optimized sequence. By adjusting the order parameter of input values based on their relevance to the classification task, the system improves classification accuracy while the determining unit manages the complexity, transforming the input data structure to match the neural network's optimal processing requirements
2Measurement precision
If input values are properly ordered to improve accuracy, then classification accuracy improves, but the processing complexity increases
Solution Approach 1:
The optimal input ordering is determined in advance before the neural network processing begins. The determining unit establishes the sequence once, and this pre-determined order is then used consistently, avoiding repeated ordering operations during processing and thus preventing excessive time loss while maintaining improved classification accuracy
Solution Approach 2:
Once the optimal input order is determined, it can be applied as a template or copy to multiple datasets. The determining unit establishes the ordering rule once, and this rule is then copied and applied to subsequent processing tasks, reducing the time required for each individual processing operation while maintaining high classification accuracy
Data Source
AI summary
A machine learning apparatus determines an order in which numerical values in an input dataset are to be entered to a neural network for data classification, based on a reference pattern that includes an array of reference values to provide a criterion for ordering the numerical values. The machine learning apparatus then calculates an output value of the neural network whose input-layer neural units respectively receive the numerical values arranged in the determined order. The machine learning apparatus further calculates an input error at the input-layer neural units, based on a difference between the calculated output value and a correct classification result indicated by a training label. The machine learning apparatus updates the reference values in the reference pattern, based on the input error at the input-layer neural units.