Neural Network Path Exploration for D-Dimensional Data Sequencing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Machine learning algorithms, particularly neural networks, are limited in processing non-sequential data as they are typically designed for static input data, such as 2D or 3D images, and struggle to effectively handle d-dimensional input data that requires sequential processing.
Innovation Solution
A computer-implemented method that explores distinct paths in d-dimensional input data to form sequences of objects, which are then processed by a neural network using machine learning algorithms, allowing for the transformation of non-sequential data into sequential data for effective processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If neural networks are designed to process static input data (2D/3D images), then they can effectively handle image data, but they cannot effectively process d-dimensional input data that requires sequential processing
Solution Approach 1:
The patent segments d-dimensional input data into sequences of objects by exploring multiple distinct paths through the data space. Each path generates a sequence that can be processed by neural networks, transforming the static data processing approach into a sequential processing framework that maintains reliability while expanding adaptability.
Solution Approach 2:
The patent introduces a sequential dimension by exploring multiple paths through d-dimensional space. This transforms static multi-dimensional data into time-ordered sequences, adding a temporal dimension that enables neural networks to process complex d-dimensional input data effectively while maintaining their architectural integrity.
2Productivity
If sequential processing algorithms are used, then they can be implemented very efficiently in hardware, but they are typically designed to process static input data and cannot handle d-dimensional input data
Solution Approach 1:
The patent creates a universal processing framework that can handle both static input data and d-dimensional input data through the same sequential processing pipeline. By transforming any d-dimensional data into sequences via path exploration, the system achieves multi-functionality that maintains hardware efficiency while expanding data type capability.
3Quantity of substance
If multiple paths are explored in d-dimensional input data, then more sequences can be formed for processing, but the complexity of the processing system increases
Solution Approach 1:
The patent applies partial action by allowing selective exploration of N distinct paths where N can be adjusted based on computational resources. This enables the system to generate sufficient sequences for effective processing without requiring exhaustive exploration of all possible paths, thereby controlling system complexity while maintaining adequate sequence quantity.
Data Source
AI summary
A method and computer program product for obtaining values are run using a neural network according to a machine learning algorithm. One embodiment may comprise accessing one or more datafiles of input data, where the input data is representable in a d-dimensional space, with d>1. The method may explore N distinct paths of the input data in the d-dimensional space, where N≥1, and collects data along the N distinct paths explored to respectively form N sequences of M objects each, with M≥2. For one or more sequences of the N sequences formed, values obtained from the M objects of each sequence may be coupled into one or more input nodes of a neural network, which is then run according to the machine learning algorithm to obtain L output values from, L≥1.