Dynamic Data Operations Modeling Using Multinomial Logistic Regression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data management systems rely on hard-coded, rules-based decision tree models that are not representative of evolving organizational policies, requiring manual updates and are inefficient in generating real-time predictions on data operations adherence to accuracy and completeness criteria.
Innovation Solution
A system for dynamic data operations modeling using a detection model based on multinomial logistic regression, which generates categorical predictions in real-time by training on meta attributes and user inputs, allowing for continuous updates and adherence assessments during data operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If hard-coded, rules-based decision tree models are used, then the system structure is simple and easy to implement, but the system cannot adapt to evolving organizational policies and requires manual updates
Solution Approach 1:
The patent transforms the static, hard-coded decision tree model into a dynamic machine learning model that continuously learns from new data. The model is retrained periodically with updated datasets, enabling it to adapt to evolving organizational policies automatically without manual intervention, thus resolving the contradiction between adaptability and complexity.
Solution Approach 2:
The system implements self-service by automatically retraining the machine learning model with new data without requiring manual updates. The automated pipeline retrieves updated data, retrains the model, and deploys new predictions, eliminating the need for manual model maintenance while maintaining adaptability to changing policies.
2Productivity
If manual updates to decision tree models are performed, then the system maintains control over model changes, but the process is time-consuming and inefficient
Solution Approach 1:
The system performs preliminary action by pre-processing and storing training data in advance, and by automating the model retraining pipeline. When new data becomes available, the system can quickly retrain the model without manual intervention, significantly reducing the time required for model updates and improving prediction generation efficiency.
Solution Approach 2:
The system implements feedback loops where prediction results and new data continuously feed back into the model training process. This automated feedback mechanism enables the system to learn from new information and update predictions efficiently, eliminating the time loss associated with manual model updates.
3Measurement precision
If traditional decision tree models are used, then the system is easy to operate, but it cannot generate real-time predictions on data operations adherence
Solution Approach 1:
The patent replaces the mechanical, rules-based decision tree system with a machine learning-based predictive system. This substitution enables the system to generate real-time predictions on data operations adherence with higher accuracy by learning patterns from historical data, while the automated pipeline maintains operational simplicity through programmatic execution.
4Reliability
If a machine learning model is trained on historical data, then the model can provide predictive insights, but the training process requires significant computational resources and time
Solution Approach 1:
The system applies partial action by training the model incrementally on subsets of data and using cross-validation techniques. This approach provides reliable predictive insights while reducing computational resource consumption compared to training on the entire dataset at once, balancing reliability with energy efficiency.
Data Source
AI summary
Systems and methods for dynamic data operations modeling. A system may include a processor and a memory storing processor-executable instructions that configure the processor to: retrieve, at time stages, a set of data records associated with meta attributes representing operations on the data records at stages over time; generate, at the respective time stages, a categorical prediction associated with the data operations based on a detection model and a set of meta attributes associated with the retrieved data records, the detection model based on a multinomial logistic regression providing the categorical prediction for adapting multi-label predictions; and transmit, following the successive time stages, one or more signals representing the categorical prediction for dynamically updating the user interface for communicating an interim categorical prediction during data operations execution.