Machine Learning Risk Scoring for Failed Computing Actions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing computing systems face inefficiencies and resource wastage due to incorrect resolution of conditional computing tasks, leading to unnecessary additional computing obligations and potential system degradation from risky transactions or actions.
Innovation Solution
Implementing a predictive machine learning model that assesses risk by incorporating expected value and using a multi-label neural network to evaluate user transactions and actions, providing risk scores over multiple time periods, thereby optimizing resource usage and reducing system burden.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a computing task makes a decision that allows the task to be performed, then the task can be completed, but it may result in future computing obligations that waste resources
Solution Approach 1:
The system performs preliminary risk assessment before allowing computing tasks to proceed. By evaluating whether a task is likely to succeed or fail in advance, the system can prevent resources from being allocated to tasks that would ultimately fail, thus avoiding future computing obligations and resource wastage.
Solution Approach 2:
The system incorporates feedback mechanisms where the outcomes of computing tasks are fed back into the risk assessment model. This allows the system to learn from past successes and failures, improving its ability to predict future task outcomes and make more accurate decisions about resource allocation.
2Loss of energy
If a computing task is denied permission, then future computing obligations are avoided, but legitimate tasks may be incorrectly rejected
Solution Approach 1:
The system uses multiple parameters and features in the risk assessment model, including historical task outcomes, task complexity, user behavior patterns, and contextual information. By considering multiple parameters simultaneously, the system achieves a balance between preventing resource wastage and maintaining high decision accuracy.
Solution Approach 2:
The risk assessment system is dynamic and adapts to changing conditions. It continuously updates its risk predictions based on new data and evolving patterns, allowing it to respond to unique situations and adjust its decisions to better reflect current realities, thereby reducing false denials.
3Loss of energy
If risk assessment is performed for all computing actions, then resource wastage is reduced, but system complexity increases
Solution Approach 1:
The system segments the risk assessment process into different components and prioritizes them based on task importance. Not all computing actions require the same level of assessment depth; the system can apply lightweight checks to low-risk tasks and more comprehensive analysis only to high-value or high-risk tasks, reducing overall system complexity.
Solution Approach 2:
The system introduces an intermediary risk assessment layer that sits between the computing task initiator and the resource allocation system. This intermediary layer filters and pre-processes risk information, reducing the complexity burden on the main computing system while still providing comprehensive risk evaluation where needed.
Data Source
AI summary
A method includes receiving training data respective of a plurality of previous computing actions by a plurality of entities and training, based on the training data, a machine learning model to output a decision score. The training includes use of a loss function having input parameters including a percentage likelihood that a future computing action will fail, and a quantitative measure of a failed computing action. The method further includes receiving, from a user entity, a request for permission to engage in further computing actions, applying the trained machine learning model to data respective of the user entity to generate a decision score respective of the user entity, and rejecting, based on the decision score, the request for permission.