Time Series Forecasting with Conformal Prediction Confidence
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional data analysis techniques for time series data-based forecasting are labor-intensive and lack quantifications of confidence and credibility values, making them inefficient for analyzing multiple data series from IoT devices and other telemetry systems.
Innovation Solution
The implementation of machine learning techniques, specifically segmented regression integrated with a greedy algorithm, and a conformal prediction framework to generate automated forecasts with confidence and credibility values for time series data from multiple devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If machine learning techniques with conformal prediction framework are applied, then forecasting automation and confidence quantification are improved, but system complexity increases
Solution Approach 1:
The patent replaces manual mechanical data analysis processes with automated machine learning algorithms. The conformal prediction framework automatically computes confidence and credibility values through algorithmic processing rather than manual statistical evaluation, enabling automated forecasting while managing system complexity through standardized computational procedures.
Solution Approach 2:
The system performs self-service by automatically generating forecasts and computing confidence metrics without requiring manual intervention. The conformal prediction framework self-calibrates to provide reliable confidence intervals and credibility assessments, reducing the need for external validation and manual quality control.
2Reliability
If conformal prediction framework is used to compute confidence values, then forecast reliability is improved, but computational time increases
Solution Approach 1:
The system performs preliminary actions by pre-computing and storing confidence and credibility values during the forecast generation process. The conformal prediction framework calculates these reliability metrics concurrently with the forecast itself rather than requiring separate validation steps, reducing overall computational time while maintaining reliability.
Solution Approach 2:
The patent maintains continuity of useful action by integrating confidence computation into the forecast generation pipeline. The conformal prediction framework continuously computes reliability metrics alongside the forecasting process, eliminating idle validation steps and ensuring that computational resources are utilized efficiently throughout the entire analysis process.
Data Source
AI summary
Methods, apparatus, and processor-readable storage media for analyzing time series data for sets of devices using machine learning techniques are provided herein. An example computer-implemented method includes processing time series data from multiple devices; generating at least one data forecast by applying, in response to a request from at least one user, one or more machine learning techniques to at least a portion of the processed time series data; computing one or more qualifying values attributable to the at least one generated data forecast by providing the at least one generated data forecast and the at least a portion of the processed time series data to a conformal prediction framework; and performing one or more automated actions based at least in part on the at least one generated data forecast and the one or more computed qualifying values.