I am using machine learning to interpret sensor signals from my phone. When I used a specific phone (Samsung), I got accurate results. However, when I used another phone (HUAWEI), it appears that the result is different.
Is this because of the different configuration in these phones? Are they using different sensors for the same type (say, accelerometer)?
If they were different, what can I do to compensate for the difference. Say, I trained my algorithm with data from Samsung, what can I do so that the algorithm will still work with HUAWEI phones accurately?
yes, Build in sensors are vary from smartphone to smartphone. The sensors quality and accuracy depends on the cost of the smartphone. But the software results are pretty much accurate. Refer the link it might be useful: https://www.researchgate.net/post/Accuracy_of_sensors_in_Android_devices