Apparatus and methods for labeling time-series data using machine learning models

US12387100B2 · kind B2 · utility

Assignee

• Anumana, Inc. · US

Inventors

• Rohit Jain · Danville, US
• Anand Ramani · Bengaluru, IN
• Yogisha H J · Bengaluru, IN
• Sanjeev Shrinivas Nadapurohit · Bengaluru, IN
• Karthik K. Bharadwaj · Bengaluru, IN
• Shiva Verma · Kanchinakote, IN

Key dates

• Filing date: Dec 8, 2024
• Grant date: Aug 12, 2025
• Expiry date: Dec 8, 2044

Classification

• Technology area (CPC G): Physics
• CPC primary: G06N3/04
• WIPO field: Computer technology
• WIPO sector: Electrical engineering

Abstract

An apparatus for labeling time-series data using machine learning models, comprising a processor and a memory containing instructions configuring the processor to receive time-series data, identify a plurality of time-series segments from received time-series data, pre-train at least a classifier using labeled training data, annotate, at a labeling module, each time-series segment of the plurality of time-series segments with at least one segment identification, retrain the at least a classifier using the annotated plurality of time-series segments, generate, using the at least a classifier, one or more segment identifications at each time-series segment subsequently identified based on continuous time-series data, and display a visual representation of the continuous time-series data with the segment identifications on a user interface.