ST provides you a lot of material to start with and it depends, basically, from the type of application you would like to develop with Machine Learning.
We can try to summarize what are the necessary basic tools. To start a Machine Learning design based on the sensor MLC, you will need to acquire some data in certain scenarios. To acquire data, you need a platform with ML sensors. There are many platforms, but we can suggest Sensortile.box for a beginner. Using ST BLE Sensor app (available for Android and iOS) you can easily start you acquisition campaign without writing a line of code. Once the dataset is available, transfer it to your PC and start UNICO-GUI software, where you can design your decision tree with an easy step-by-step process. Once the design is completed, you just need to transfer the configuration file back to your phone, program the sensor with the app and you will have a ML application running and ready to be tested.
You can find many detailed tutorials on our GitHub page dedicated to MLC; in particular, you can find a more detailed guide of a Yoga Position application enabled by Sensortile.box and STBLE App HERE. This example can be easily converted in any application based on ML.
Finally, if you are interested more in the process of generating a decision tree and you don’t know much about feature selection and other tricky topics related to it, you can either read a detailed document HERE or watch a series of 2 YouTube videos where our colleague summarizes all the topics that you should know to get into the Decision Tree (and ML) universe.
