18315487. MACHINE LEARNING AUTOMATED SIGNAL DISCOVERY FOR FORECASTING TIME SERIES simplified abstract (INTERNATIONAL BUSINESS MACHINES CORPORATION)

From WikiPatents
Jump to navigation Jump to search

MACHINE LEARNING AUTOMATED SIGNAL DISCOVERY FOR FORECASTING TIME SERIES

Organization Name

INTERNATIONAL BUSINESS MACHINES CORPORATION

Inventor(s)

Xuan-Hong Dang of Chappaqua NY (US)

Petros Zerfos of New York NY (US)

Syed Yousaf Shah of Yorktown Heights NY (US)

Anil R. Shankar of New York NY (US)

MACHINE LEARNING AUTOMATED SIGNAL DISCOVERY FOR FORECASTING TIME SERIES - A simplified explanation of the abstract

This abstract first appeared for US patent application 18315487 titled 'MACHINE LEARNING AUTOMATED SIGNAL DISCOVERY FOR FORECASTING TIME SERIES

The abstract of the patent application describes a prediction system that uses machine learning models to analyze data from devices and forecast time series.

  • Simplified Explanation: The patent application outlines a system that uses machine learning to analyze data from devices and predict future trends.
  • Key Features and Innovation:

- Utilizes a first machine learning model to identify signals from data. - Trains a second machine learning model to forecast a time series and evaluate its performance. - Refines signals if performance does not meet a threshold and trains a third machine learning model for improved forecasting. - Predicts the performance of a third forecasted time series using refined signals and the third model.

  • Potential Applications:

- Weather forecasting - Stock market predictions - Traffic flow analysis - Energy consumption forecasting

  • Problems Solved:

- Improves accuracy of forecasting models - Enhances performance evaluation of predictions - Allows for iterative refinement of signals for better forecasting results

  • Benefits:

- Increased accuracy in forecasting - Better decision-making based on predictions - Enhanced efficiency in resource allocation

  • Commercial Applications:

- Financial institutions for market analysis - Energy companies for demand forecasting - Transportation agencies for traffic management

  • Prior Art:

- Researchers in the field of machine learning and predictive analytics - Companies specializing in data analysis and forecasting technologies

  • Frequently Updated Research:

- Ongoing advancements in machine learning algorithms for time series forecasting

Questions about the technology: 1. How does the prediction system refine signals to improve forecasting accuracy? 2. What are the potential limitations of using machine learning for time series forecasting?


Original Abstract Submitted

A prediction system may obtain data, via a network, from devices and process the data, using a first machine learning, to identify a plurality of signals. The prediction system may train a second machine learning model to analyze the plurality of signals to forecast a first forecasted time series and evaluate a first performance of the first forecasted time series. The prediction system may determine that the first performance does not satisfy a performance threshold and may refine the plurality of signals to obtain a refined plurality of signals. The prediction system may train a third machine learning model to analyze the refined plurality of signals to forecast a second forecasted time series and evaluate a second performance of the second forecasted time series. The prediction system may use the refined plurality of signals and the third machine learning model to predict a performance of a third forecasted time series.

Retrieved from "http://wikipatents.org/index.php?title=18315487._MACHINE_LEARNING_AUTOMATED_SIGNAL_DISCOVERY_FOR_FORECASTING_TIME_SERIES_simplified_abstract_(INTERNATIONAL_BUSINESS_MACHINES_CORPORATION)&oldid=86177"