Time Series Annotation Using Adaptive Clustering Validation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The scarcity of labelled data for real-time datasets poses a challenge, leading to high resource and time costs for data annotation, and the analyzed data may be prone to errors, especially in supervised machine learning models.
Innovation Solution
A processor-implemented method for annotating and validating time series data using edge devices, which applies clustering techniques, transmits subsets to the cloud for validation, and iteratively refines error rates using meta data and user inputs to generate machine learning models.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If data annotation is performed manually to obtain labelled data for supervised machine learning, then model performance is improved, but resource cost and time consumption increase significantly
Solution Approach 1:
The system performs self-annotation by automatically generating cluster labels through unsupervised clustering algorithms. The edge device clusters time series data and assigns labels without human intervention, while the cloud validates these labels. This self-service mechanism eliminates the need for manual annotation, reducing time and resource costs while maintaining model performance through automated label generation and validation.
2Reliability
If all time series data is annotated and validated to ensure high accuracy, then model reliability is improved, but processing time and computational resources increase
Solution Approach 1:
The system segments the annotation and validation process into two distinct parts: unsupervised clustering performed locally at the edge device, and validation performed selectively at the cloud. This segmentation allows the majority of data to be processed quickly at the edge without cloud validation, while only selected portions require cloud-based validation, thereby maintaining reliability without sacrificing processing efficiency.
Solution Approach 2:
Instead of validating all annotated data, the system applies partial validation only to selected clusters or samples at the cloud. This partial action approach ensures that model reliability is maintained through validation of critical data portions, while avoiding the excessive computational overhead of validating every single data point, thus preserving productivity.
3Measurement precision
If manual data annotation is performed to reduce errors in analyzed data, then data quality is improved, but cost in terms of resources and time increases
Solution Approach 1:
The system replaces the mechanical process of manual human annotation with automated computational processes. Unsupervised clustering algorithms automatically assign labels to time series data, and cloud-based validation mechanisms automatically verify these labels. This substitution eliminates the need for human resources while maintaining data quality through algorithmic label generation and validation, thereby reducing resource costs without compromising precision.
Data Source
AI summary
Conventionally, applying analytics on dataset is the scarcity of labelled data. With increase of data there is cost fact effecting nature of servicing required for data (e.g., cost in terms of resource and time and effort is high for data annotation). Though data is analysed, it may be prone to error. Present disclosure provides systems/methods for reducing volume of data to be annotated for time series data thereby reducing time and effort of resources, thus resulting in effective utilization of system's resources (e.g., memory, processor, etc.). More specifically, the method of the present disclosure adaptively modifies the volume of the data to be annotated based on the performance of the unsupervised learning method applied in the system. Moreover, in the absence of an annotation mechanism for clusters of time series data, meta data associated with the time series data is utilized for annotation and validation of dataset.