Neural Network Expense Report Risk Determination System
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The growing volume of expense reports leads to increased latency and fraud in reimbursement processes due to the inefficiencies of traditional expense report determination systems, which are overwhelmed by the volume and variety of data, and are incompatible with conventional machine training techniques.
Innovation Solution
A computerized system that balances imbalanced datasets of expense reports by creating a subsample and using neural network learning models to extract feature data, train models, and determine risk scores for expense reports, enabling automated approval, rejection, or flagging without user input.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional expense report determination systems are used to process growing volume of expense reports, then processing capacity is maintained with conventional methods, but processing latency increases and fraud detection accuracy deteriorates
Solution Approach 1:
The patent replaces conventional manual expense report determination systems with a neural network-based machine learning system. The neural network model automatically processes expense reports by learning patterns from training data, substituting manual review processes with automated intelligent analysis that can handle high volumes without increasing latency.
Solution Approach 2:
The system implements self-service through the neural network's ability to autonomously determine expense report outcomes without human intervention. The model independently processes reports, applies learned criteria, and generates determinations, enabling the system to serve itself in handling the growing volume of expense reports efficiently.
2Ease of manufacture
If conventional machine training techniques are used on typical expense reports, then training process simplicity is maintained, but training compatibility fails due to data incompatibility
Solution Approach 1:
The patent transforms expense report data into a format compatible with neural network training by changing data parameters and structure. The system converts unstructured or semi-structured expense report data into structured numerical representations that the neural network can process, making the training data compatible while maintaining the essential information needed for accurate determinations.
Solution Approach 2:
The patent introduces data preprocessing and feature extraction as intermediary steps between the raw expense reports and the neural network training process. These intermediary components transform the incompatible data format into a suitable representation, acting as a bridge that enables conventional training techniques to work effectively with expense report data.
3Productivity
If traditional expense report systems process high volume reports, then processing throughput is maintained, but fraud detection accuracy deteriorates due to system overload
Solution Approach 1:
The patent replaces traditional rule-based fraud detection mechanisms with a neural network system that learns complex patterns from training data. This substitution enables the system to maintain high fraud detection accuracy even when processing high volumes of reports, as the neural network can identify subtle fraudulent patterns that traditional systems miss under load.
Solution Approach 2:
The patent implements preliminary training of the neural network on labeled expense report data before deployment. This preliminary action allows the system to learn fraud patterns and determination criteria in advance, so that during high-volume processing, the pre-trained model can quickly and accurately detect fraud without being overwhelmed by the volume of incoming reports.
Data Source
AI summary
Aspects of the disclosure provide a computerized method and system that utilizes reference expense reports to build and train one or more neural network learning models that intelligently determine the riskiness of to-be-determined expense reports submitted for reimbursement. In examples, a determined riskiness may inform a reimbursement entity manager when determining whether to approve, reject, and/or flag for further review a to-be-determined expense report. In instances, computerized expense report resolution systems and methods may be further automated in order to omit user interactions with to-be-determined expense reports, such that an intelligent computer determines whether to approve, reject, and/or flag a to-be-determined expense report based on the intelligently determined riskiness of the to-be-determined expense report.