TDK Corporation (20240311629). ONLINE LEARNING METHOD AND ONLINE LEARNING DEVICE simplified abstract
Contents
ONLINE LEARNING METHOD AND ONLINE LEARNING DEVICE
Organization Name
Inventor(s)
ONLINE LEARNING METHOD AND ONLINE LEARNING DEVICE - A simplified explanation of the abstract
This abstract first appeared for US patent application 20240311629 titled 'ONLINE LEARNING METHOD AND ONLINE LEARNING DEVICE
Simplified Explanation: The patent application describes an online learning method that involves compressing and expanding the range of possible values of a Kalman gain before and after an update, respectively.
- **Key Features and Innovation:**
* Compressing the range of possible values of a Kalman gain before an update. * Obtaining a Kalman gain after the update using an expanded Kalman filter method. * Expanding the range of possible values of the Kalman gain after the update. * Updating a weight based on the error between a training signal and an inference result.
- **Potential Applications:**
* Online learning algorithms. * Predictive modeling. * Signal processing.
- **Problems Solved:**
* Efficiently updating Kalman gains in online learning scenarios. * Improving the accuracy of inference results.
- **Benefits:**
* Enhanced learning performance. * Improved prediction accuracy. * Reduced computational complexity.
- **Commercial Applications:**
* Predictive analytics software. * Financial forecasting tools. * Autonomous vehicle systems.
- **Prior Art:**
Prior art related to this technology may include research on Kalman filtering methods in online learning and signal processing.
- **Frequently Updated Research:**
Ongoing research in the field of online learning algorithms and Kalman filtering techniques may provide further insights into the optimization of this method.
Questions about Online Learning Method: 1. How does the compressed Kalman gain before an update impact the efficiency of the online learning process? 2. What are the potential implications of expanding the range of possible values of the Kalman gain after an update on predictive modeling accuracy?
Original Abstract Submitted
an online learning method includes: compressing a range of possible values of a kalman gain before an update; obtaining a kalman gain after the update from the compressed kalman gain before the update using an expanded kalman filter method; expanding the range of possible values of the kalman gain after the update, and updating a weight by adding a weight before the update to a result obtained by multiplying the kalman gain in which the range of the possible values of the kalman gain is expanded by an error between a training signal and an inference result in which a weight before the update is used.