17806889. MACHINE LEARNING APPROACH FOR SOLVING THE COLD START PROBLEM IN STATEFUL MODELS simplified abstract (Microsoft Technology Licensing, LLC)
MACHINE LEARNING APPROACH FOR SOLVING THE COLD START PROBLEM IN STATEFUL MODELS
Organization Name
Microsoft Technology Licensing, LLC
Inventor(s)
Andrey Karpovsky of Kiryat Motzkin (IL)
MACHINE LEARNING APPROACH FOR SOLVING THE COLD START PROBLEM IN STATEFUL MODELS - A simplified explanation of the abstract
This abstract first appeared for US patent application 17806889 titled 'MACHINE LEARNING APPROACH FOR SOLVING THE COLD START PROBLEM IN STATEFUL MODELS
Simplified Explanation
The patent application describes a computing system that generates an initial profile based on user input and extracts features from received datasets to determine if their behavioral patterns are anomalous. The system includes both first and second machine-learning models trained on subsets of the received datasets.
- The computing system generates an initial profile based on user input.
- The initial profile specifies expected behavioral patterns of received datasets.
- Features indicative of behavioral patterns are extracted from the received datasets.
- The initial profile is provided to first machine-learning models.
- The first machine-learning models determine if the received datasets have anomalous behavioral patterns.
- The system also includes second machine-learning models trained on subsets of the received datasets.
- The second machine-learning models train a second profile based on the extracted features.
- The second profile specifies the behavioral patterns learned by the second machine-learning models.
Potential Applications
This technology can have various applications in fields such as:
- Cybersecurity: Detecting anomalous patterns in network traffic or user behavior.
- Fraud detection: Identifying unusual patterns in financial transactions.
- Predictive maintenance: Monitoring equipment behavior to detect potential failures.
- Quality control: Identifying deviations in manufacturing processes.
Problems Solved
The technology addresses the following problems:
- Identifying anomalous patterns in large datasets can be time-consuming and challenging for humans.
- Traditional rule-based systems may not capture all possible anomalies.
- Training machine-learning models on subsets of datasets allows for more accurate anomaly detection.
- The system provides a way to continuously learn and update behavioral patterns based on new data.
Benefits
The technology offers several benefits:
- Improved accuracy in detecting anomalous patterns in datasets.
- Reduction in false positives and false negatives compared to traditional methods.
- Ability to adapt and learn from new data to improve anomaly detection over time.
- Automation of the process, saving time and resources for manual analysis.
Original Abstract Submitted
A computing system generates from received user input an initial profile. The initial profile specifies expected behavioral patterns of datasets that are to be received by the computing system. The computing system extracts from received datasets features that are indicative of behavioral patterns of the received datasets. The computing system provides the initial profile to first machine-learning models. The first machine-learning models have been trained using a subset of the received datasets. The first machine-learning models use the initial profile to determine if the behavioral patterns of the received datasets are anomalous. The computing system includes second machine-learning models that have been trained using a subset of the received datasets. The second machine-learning models train a second profile based on the extracted features to specify behavioral patterns of the received datasets that are learned by the second machine-learning model.
