Quality Over Quantity - Active Learning Behind the Scenes
from Healthy.io (Israel)
About speaker
Head of Research, Healthy.io
Sivan Biham is a Machine Learning researcher with experience in Computer Vision, Reinforcement Learning and Deep Learning among others. Currently, she is the Head of Research at Healthy.
About speakers company
Using colourimetric analysis, computer vision, and AI we transform the smartphone camera into a clinical-grade medical device. Our computer vision algorithms and unique calibration method can analyse urine tests and chronic wounds using a variety of devices and under different lighting conditions, making testing as easy as snapping a picture.
Abstracts
In this talk, Sivan will show how you can train machine learning models with less data by focusing on quality over quantity. Through the power of Active Learning, she’ll explain how to collect the most valuable data to save time and resources while still achieving top model performance. If you want to learn how to get better results with fewer labeled samples, this is the talk for you!
In the world of machine learning, more data doesn't always mean better models. The key is not just quantity but quality. "Quality Over Quantity - Active Learning Behind the Scenes" explores the concept of Active Learning, a powerful approach for optimizing the data collection process. In this talk, Sivan will clarify how focusing on the most valuable data can help achieve high-performing models with fewer labeled samples, saving both time and resources. She will share insights from her work in Active Learning, including the methods she used to prioritize data collection, the research process, and how to validate the results. This talk will provide practical tips and strategies for applying Active Learning to your own projects, ensuring that your models are both efficient and effective. Whether you're a researcher or practitioner, you’ll leave with a fresh perspective on data collection and how to get the most from your labeled datasets.
The Program Committee has not yet taken a decision on this talk
other talks of this topic
Fabien Vauchelles
Scrapoxy
Shelly Goldblit
Dell Technologies
Vadzim Prudnikau
Trainitek
Daniel Raniz Raneland
factor10
Opemipo Disu
Latitude
Daniel Raniz Raneland
factor10
Alexandre Gallice
Red Hat
Federico Fregosi
OpsGuru
João Esperancinha
Vereniging COIN
Ambesh Singh
Visionet Systems Deutschland
Vadzim Prudnikau
Trainitek
Katharina Fetzer
hylane GmbH
Sivan Biham
Healthy.io
Pradeep Sharma
Independent
Ambesh Singh
Visionet Systems Deutschland
Siddhant Agarwal
Neo4j
Tech Internals Conf is the leading conference for developers of complex and highly loaded systems
Participation options
Offline
The price is soaring —> the closer the conference is, the more it costs.
The current price of a ticket is —> 360 EUR
If you have any questions you can reach out to our support service —> support@internals.tech
Special offer (from 5 tickets)
To order from 5 tickets, contact us support@internals.tech
leave a request
Changed your mind?
Please tell us why.
Thank you for your reply!
Become partner
Professional conference for developers of high-load systems