View source for 18300340. SYSTEMS AND METHODS FOR OUTLIER DETECTION USING UNSUPERVISED MACHINE LEARNING MODELS TRAINED ON BALANCED DATA simplified abstract (Capital One Services, LLC)

=SYSTEMS AND METHODS FOR OUTLIER DETECTION USING UNSUPERVISED MACHINE LEARNING MODELS TRAINED ON BALANCED DATA=


==Organization Name==

[[:Category:Capital One Services, LLC|Capital One Services, LLC]]


[[Category:Capital One Services, LLC]]

==Inventor(s)==

[[:Category:Hassan Shallal of Plano TX (US)|Hassan Shallal of Plano TX (US)]][[Category:Hassan Shallal of Plano TX (US)]]

[[:Category:Rajesh Kanna Durairaj of Plano TX (US)|Rajesh Kanna Durairaj of Plano TX (US)]][[Category:Rajesh Kanna Durairaj of Plano TX (US)]]

==SYSTEMS AND METHODS FOR OUTLIER DETECTION USING UNSUPERVISED MACHINE LEARNING MODELS TRAINED ON BALANCED DATA - A simplified explanation of the abstract==


This abstract first appeared for US patent application 18300340 titled 'SYSTEMS AND METHODS FOR OUTLIER DETECTION USING UNSUPERVISED MACHINE LEARNING MODELS TRAINED ON BALANCED DATA

The abstract describes methods and systems for outlier detection, utilizing either a single-tier model with a minority expert or a two-tier model with both a majority expert and a minority expert.

* The system may create the minority expert model by training an unsupervised machine learning model on oversampled training data, which includes synthetic sample outlier events.
* The minority expert model can provide a binary result indicating if an event is an outlier or not, or a multi-class result indicating sub-categories of the outlier category.
* The system can perform outlier detection on sequence-based or non-sequence-based events.

Potential Applications:
- Fraud detection in financial transactions
- Anomaly detection in network traffic
- Quality control in manufacturing processes

Problems Solved:
- Identifying outliers in large datasets
- Improving accuracy in anomaly detection
- Enhancing security measures against fraudulent activities

Benefits:
- Increased efficiency in outlier detection
- Reduction in false positives
- Enhanced data security and integrity

Commercial Applications:
Title: Advanced Outlier Detection System for Enhanced Security Measures
This technology can be applied in industries such as finance, cybersecurity, and manufacturing to improve anomaly detection and enhance overall security measures.

Prior Art:
Further research can be conducted in the field of machine learning models for outlier detection, particularly focusing on the use of minority expert models in two-tier systems.

Frequently Updated Research:
Stay updated on the latest advancements in machine learning algorithms for outlier detection, as well as new techniques for training expert models in anomaly detection systems.

Questions about Outlier Detection:
1. How does the two-tier machine learning model improve outlier detection compared to a single-tier model?
2. What are the potential challenges in implementing synthetic sample outlier events in training data for the minority expert model?


==Original Abstract Submitted==

Methods and systems are described herein for outlier detection. The system may apply a single-tier including a minority expert model only or a two-tier machine learning model, including a majority expert model and a minority expert model. The system may generate the minority expert model by training an unsupervised machine learning model on oversampled training data, including synthetic sample outlier events. In some embodiments, the minority expert model may provide a binary result indicating an event belongs to an outlier category or not. In some embodiments, the minority expert model may include multiple component models providing a multi-class result indicating whether an event belongs to a sub-category of the outlier category. In application, the system may perform the outlier detection on events that are sequence-based or not.



[[Category:G06Q20/40]]
[[Category:CPC_G06Q20/4016]]