[Original article published by Tan Chin Luh – “When Machine Learning Meets IoT“]
In our upcoming training “IoT Made Easy with Favoriot and Scilab IoT Module“, we will be building an IoT system prototype by using a simple Machine Learning example.
Loading Data Set
The dataset consists of 2 sensors reading, one from the light sensor, another from the voltage divider with a variable resistor.
The idea is pretty simple: Turning on LEDs or sound the buzzer when the 2 sensors are falling in the region of the “blue dots”.
Training The Neural Network And Visualizing the Decision Boundary
The data is feed into neural network provided in Scilab Neural Network Module and trained for classification purpose. The result is shown as below:
Testing with Arduino and Gadgets
The simple system as below is built and interface with a computer to test the neural network in real-time.
The video below illustrates the result when the sensors reading (red circle) reach the “blue dot” region, the buzzer will beep and the LEDs would be turned on, and off when it reaches the “green star” region.
Trying With Data from MQTT Server
Finally, we will try it out with the signal from the MQTT server!
Join Us At:
IoT Made Easy with Favoriot and Scilab IoT Module Training
FAVORIOT offered a very affordable pricing plan as shown below:
The Beginner Plan is very suitable for University students who have been assigned or chose IoT project as their Final Year Project. A single device such as Raspberry Pi or Arduino that collects from several connected sensors can stream to the IoT platform. A simple dashboard is provided to visualize the measurements. This plan is also suitable for an individual developer or hobbyist that would like to familiarize with IoT platform, conduct self-learning or participate in IoT Hackathons. At RM 8.33 (or less than USD 2.00) per month, you can become the next Generation-IoT and on your way to be one of the rare talents in the job market today!
The Startup Plan is perfect for Freelancers or SMEs or Startups to provide IoT solution for a medium size project up to 150 devices. In fact, they can also create multiple projects or applications but limited to the total number of 150 devices.
The Business Plan is great for developing big commercial IoT project. It can support up to 300 devices for many smart applications.
However, if a company requires connecting thousands of devices, they may contact/email their customized requirements to sales@favoriot.com
CLICK HERE to Subscribe/Request for Quotation.
Check out the COMPLETE Tutorial How to connect an Arduino to FAVORIOT Platform. You can also download a STEP-by-STEP Tutorial – TUTORIAL – Cytron Uno with FAVORIOT Platform (53 downloads)