Yes, we suggest to adopt Unico GUI because all the MLC process is easy with just one tooI, but you can approach the Machine Learning process in other ways.
For example, the decision tree can be built also using external tools such as Weka, Python, RapidMiner and MATLAB. The output of these external tool is always compatible with Unico (for Python and MATLAB we provide some script as examples in our GitHub repository: https://github.com/STMicroelectronics/STMems_Machine_Learning_Core/tree/master/tools).
All these external tools are also described in the appendix sections of the MLC application notes of our devices.
Using the SensorTile.box, in the first step of the MLC process, data can be captured through the "ST BLE Sensor" app and stored into the SD card. Then data logs can be dowloaded on the PC, and the MLC configuration process can be implemented using Unico GUI.
