Running a Boosted Trees Classifier or a SVM Classifier on an Arduino

Go To Last Post
4 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi

Currently I am using a NodeMCU to process Gyroscope data by sending the data to my server using a REST endpoint which returns either a 0 or a 1

How can I run this classification on the Arduino itself, I created the initial ML Model using Turicreate, and I can even extract the coefficients for the SVM Classifier

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Wow -

 

Boosted Trees Classifier, NodeMCU, REST, ML Model, Turicreate, SVM Classifier.

 

You will have to provide a lot more information and some details about these names and acronyms before anyone can help.

 

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

Last Edited: Mon. Jan 20, 2020 - 08:36 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

ChristianHunter wrote:
How can I run this classification on the Arduino itself,
The problem with any kind of "machine learning" is that it necessarily implies a large amount of data being built up to create the knowledge base. AVR micros with just 2K SRAM or whatever do not really lend themselves to such a process. About the only hope wold be if you could add some external storage perhaps.

 

A more resourced system (like a RaspberryPi) would make a much better host for this kind of thing.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

NodeMCU is probably a ESP8266 or ESP32 based device, so it has a reasonably performant 32bit cpu, 1M+ of external flash and a good slab of ram. These can run Arduino. Maybe something like OpenML might be more suitable.