Products
Scale 3D Sensor FusionAdvanced annotations for LiDAR + RADAR data
Scale ImageComprehensive annotations for images
Scale VideoScalable annotations for video data
Scale TextSophisticated annotations for text-based data
Scale AudioAudio Annotation and Speech Annotation for NLP
Scale MappingThe flexible solution to develop your own maps
Scale NucleusThe mission control for your data
Scale Document AITemplate-free ML document processing
Scale CollectCollect diverse data
Scale SyntheticGenerate synthetic data
Solutions
CustomersPricing
Resources
Company
Hosted by Scale
The Learning Curve: Humans-in-the-Loop Don't Perform Gradient Descent
Recorded Aug 28th @ 12AM Coordinated Universal Time
About this event
The Learning Curve: Humans-in-the-Loop Don't Perform Gradient Descent
Join us for a Tech Talk!
Data labeling at scale is a complicated endeavor involving operations, machine learning and game theory. In this talk we'll focus on the game theory, and introduce a model of human-in-the-loop labeling as an optimization problem. Based on our (Scale AI's) experience of operating tens of thousands of labelers on high complexity tasks, we will discuss how various problems predicted by the model map to observed results.
We welcome all Product Managers, Machine Learning Engineers, Software Engineers, Technical Program Managers, Research Scientists, and Applied Scientists for this technical discussion about optimizing human-in-the-loop labelling.
Agenda
- Intro to Human-in-the-Loop Labeling
- Similarities to training an ML Model
- Data augmentations for training labelers
- The right loss function
- Q&A
Presented by
Akshat Bubna
Akshat is currently an engineering manager at Scale AI, where he leads a team that is building software and systems to label text, audio and computer vision ground truth data for ML teams with incredible accuracy. He previously has worked at D. E. Shaw and database and fintech startups, and studied computer scienc
...