Firstly, welcome to Shepherd! We hope that you are enjoying our services.
As part of the initial on-boarding, one of our team members would have started to set up 'machine learning' on sensors, within your site. This is crucial, as it enables our system to 'learn what normal looks like'.
Shepherd's machine learning will often make initial judgements based on the historical data that it has. This is great for speedy alerting, but may not be 100% accurate if there is new information to be taken into consideration. Often, in the first few weeks of using our service, you may get false alerts whilst the system builds its profile of what your normal operating pattern looks like.
Fixing false alerts for the future is simple.
Click 'Do not send alerts for this event in the future'. What you are doing here is 'teaching' the machine learning algorithm. It will now remember that an event like this is ok, and add it to the profile of 'what normal looks like'. Although most systems go along without a hitch, every system is different, and Shepherd needs to learn. Doing this strengthens the analytics and will lead to more accurate alerts in the future.
Thanks for reading!