tinyML Asia Video Poster Neuton AI: Bringing Big Ideas into Tiny Devices Bottoms-up Approach to…

tinyML Asia
Video Poster
Bringing Big Ideas into Tiny Devices Bottoms-up Approach to Building Extremely Small Models from Neuton.ai
Blair NEWMAN, CTO, Neuton.ai

Bringing intelligence to edge devices will measure the world in new ways, providing opportunities for making smarter data-driven decisions that can change human lives for the better.

Why is our world not there yet? Why do we still face the difficulty of embedding large ML models into edge devices and evaluating model quality? Why do traditional data science algorithms fail for TinyML?

To highlight the matter from different angles, we’ll tackle the following points:

Explain why models built with traditional frameworks are not optimal in size and accuracy.
Compare traditional mathematical algorithms with a novel approach to building compact self-organizing neural networks with an excellent generalization capability.
Demonstrate this novel approach in action and explain how even non-data scientists can build predictive models in a few clicks, without any coding or loss of accuracy.
Compare the output metrics of our model with the model build with a traditional approach by using TensorFlow Lite.
At Neuton.ai, we believe that users of any tech level should be able to get actionable insights from their data effortlessly. That’s why we strive to make the process of solving real-world challenges by machine learning, super easy and intuitive.

source