Passive data collection and use of machine-learning models for event prediction

1. A computer-implemented method comprising:

determining that data is to be monitored at a particular electronic device for indications of an event; and

in response to the determination:

accessing one or more data sets representative of one or more inputs or sensor measurements collected from one or more input devices or sensors of the particular electronic device, wherein the one or more data sets comprise communication data indicating communication interaction of a user associated with the particular electronic device with other users;

processing the one or more data sets using a trained machine-learning model to generate a result corresponding to a predicted risk of the event;

transmitting the result to a remote computing device;

reconfiguring, based on an instruction received from the remote computing device, the particular electronic device to change an operation of the particular electronic device for collecting the inputs or sensor measurements;

collecting additional input data or sensor measurements from the one or more input devices or sensors; and

processing the additional input data or sensor measurements using the trained machine-learning model to generate an updated result corresponding to an updated risk of the event.

2. The computer-implemented method of claim 1 , wherein the one or more data sets includes location data that identifies one or more locations at which the particular electronic device was located, and wherein processing the one or more data sets comprises:

identifying one or more base location areas within which the particular electronic device has been frequently located;

determining, using the location data, a time variable indicating for how long the particular electronic device was outside of the one or more base location areas; and

inputting an input data set that includes the time variable to the trained machine-learning model.

3. The computer-implemented method of claim 1 , wherein the communication data comprise data pertaining to one or more of phone calls, messages, or a de-identified contact list, and wherein processing the one or more data sets comprises:

determining, using the communication data in the one or more data sets, a communication statistic that characterizes recent call history or recent message history associated with the particular electronic device; and

inputting an input data set that includes the communication statistic to the trained machine-learning model.

4. The computer-implemented method of claim 3 , wherein the communication statistic comprises one or more of:

a count of one or more of the phone calls or outgoing messages;

a duration statistic of the one or more calls;

a length statistic of the one or more outgoing messages; or

a contact diversity statistic relating to a number of different contacts with which the one or more calls or outgoing messages were communicating.

5. The computer-implemented method of claim 1 , wherein the sensor measurements characterize movement of the particular electronic device, and wherein processing the one or more data sets comprises:

generating, based on the sensor measurements, a movement statistic representative of one or more of a duration, an intensity or a frequency of movement of the particular electronic device; and

inputting an input data set that includes the movement statistic to the trained machine-learning model.

6. The computer-implemented method of claim 1 , further comprising:

determining that an alert condition is satisfied based on the result; and

as a result of determining that the alert condition is satisfied, transmitting the result to another electronic device.

7. The computer-implemented method of claim 1 , further comprising:

accessing, by another electronic device, a user classification model based at least in part on health data for a plurality of users;

wherein determining that data is to be monitored at the particular electronic device for predictions of the event comprises:

classifying, by the other electronic device, a user into a user group using the user classification model; and

selecting, by the other electronic device, the trained machine-learning model based at least in part on the user group to which the user is classified.

8. The computer-implemented method of claim 7 , further comprising:

executing, by the other electronic device, a function on at least a portion of the health data of the plurality of users to produce group expressions;

mapping, by the other electronic device, the group expressions to an N-dimensional space; and

classifying, by the other electronic device, a region bounded by the group expressions as a user group.

9. The computer-implemented method of claim 1 , wherein the trained machine-learning model is trained a machine-learning model with health record data of a user associated with the particular electronic device.

10. A computer-program product tangibly embodied in a non-transitory machine-readable storage medium, including instructions configured to cause one or more data processors to perform operations including:

determining that data is to be monitored at a particular electronic device for indications of an event; and

in response to the determination:

accessing one or more data sets representative of one or more inputs or sensor measurements collected from one or more input devices or sensors of the particular electronic device, wherein the one or more data sets comprise communication data indicating communication interaction of a user associated with the particular electronic device with other users;

processing the one or more data sets using a trained machine-learning model to generate a result corresponding to a predicted risk of the event;

transmitting the result to a remote computing device;

reconfiguring, based on an instruction received from the remote computing device, the particular electronic device to change an operation of the particular electronic device for collecting the inputs or sensor measurements;

collecting additional input data or sensor measurements from the one or more input devices or sensors; and

processing the additional input data or sensor measurements using the trained machine-learning model to generate an updated result corresponding to an updated risk of the event.

11. The computer-program product of claim 10 , wherein the one or more data sets includes location data that identifies one or more locations at which the particular electronic device was located, and wherein processing the one or more data sets comprises:

identifying one or more base location areas within which the particular electronic device has been frequently located;

determining, using the location data, a time variable indicating for how long the particular electronic device was outside of the one or more base location areas; and

inputting an input data set that includes the time variable to the trained machine-learning model.

12. The computer-program product of claim 10 , wherein the communication data comprise data pertaining to one or more of phone calls, text messages, or a de-identified contact list, and wherein processing the one or more data sets comprises:

determining, using the communication data in the one or more data sets, a communication statistic that characterizes recent call history or recent message history associated with the particular electronic device; and

inputting an input data set that includes the communication statistic to the trained machine-learning model.

13. The computer-program product of claim 12 , wherein the communication statistic comprises one or more of:

a count of one or more of the phone calls or outgoing messages;

a duration statistic of the one or more calls;

a length statistic of the one or more outgoing messages; or

a contact diversity statistic relating to a number of different contacts with which the one or more calls or the one or more outgoing messages were communicating.

14. The computer-program product of claim 10 , wherein the sensor measurements characterize movement of the particular electronic device, and wherein processing the one or more data sets comprises:

generating, based on the sensor measurements, a movement statistic representative of one or more of a duration, an intensity, or a frequency of movement of the particular electronic device; and

inputting an input data set that includes the movement statistic to the trained machine-learning model.

15. The computer-program product of claim 10 , wherein the instructions further cause the one or more data processors to perform operations including:

determining that an alert condition is satisfied based on the result; and

as a result of determining that the alert condition is satisfied, transmitting the result to another electronic device.

16. The computer-program product of claim 10 , wherein the instructions further cause the one or more data processors to perform operations including:

accessing, by another electronic device, a user classification model based at least in part on health data for a plurality of users;

wherein determining that data is to be monitored at the particular electronic device for predictions of the event comprises:

classifying, by the other electronic device, a user into a user group using the user classification model; and

selecting, by the other electronic device, the trained machine-learning model based at least in part on the user group to which the user is classified.

17. The computer-program product of claim 16 , wherein the instructions further cause the one or more data processors to perform operations including:

executing, by the other electronic device, a function on at least a portion of the health data of the plurality of users to produce group expressions;

mapping, by the other electronic device, the group expressions to an N-dimensional space; and

classifying, by the other electronic device, a region bounded by the group expressions as a user group.

18. The computer-program product of claim 10 , wherein the trained machine-learning model is trained a machine-learning model with health record data of a user associated with the particular electronic device.

19. A system comprising:

one or more data processors; and

a non-transitory computer readable storage medium containing instructions which when executed on the one or more data processors, cause the one or more data processors to perform operations including:

determining that data is to be monitored at a particular electronic device for indications of an event; and

in response to the determination:

accessing one or more data sets representative of one or more inputs or sensor measurements collected from one or more input devices or sensors of the particular electronic device, wherein the one or more data sets comprise communication data indicating communication interaction of a user associated with the particular electronic device with other users;

processing the one or more data sets using a trained machine-learning model to generate a result corresponding to a predicted risk of the event;

transmitting the result to a remote computing device;

reconfiguring, based on an instruction received from the remote computing device, the particular electronic device to change an operation of the particular electronic device for collecting the inputs or sensor measurements;

collecting additional input data or sensor measurements from the one or more input devices or sensors; and

processing the additional input data or sensor measurements using the trained machine-learning model to generate an updated result corresponding to an updated risk of the event.

20. The system of claim 19 , wherein the one or more data sets includes location data that identifies one or more locations at which the particular electronic device was located, and wherein processing the one or more data sets comprises:

identifying one or more base location areas within which the particular electronic device has been frequently located;

determining, using the location data, a time variable indicating for how long the particular electronic device was outside of the one or more base location areas; and

inputting an input data set that includes the time variable to the trained machine-learning model.

21. The system of claim 19 , wherein the communication data comprise data pertaining to one or more of phone calls, text messages, or a de-identified contact list, and wherein processing the one or more data sets comprises:

determining, using the communication data in the one or more data sets, a communication statistic that characterizes recent call history or recent message history associated with the particular electronic device; and

inputting an input data set that includes the communication statistic to the trained machine-learning model.

22. The system of claim 21 , wherein the communication statistic comprises one or more of:

a count of one or more of the phone calls or outgoing messages;

a duration statistic of the one or more calls;

a length statistic of the one or more outgoing messages; or

a contact diversity statistic relating to a number of different contacts with which the one or more calls or the one or more outgoing messages were communicating.

23. The system of claim 19 , wherein the sensor measurements characterize movement of the particular electronic device, and wherein processing the one or more data sets comprises:

generating, based on the sensor measurements, a movement statistic representative of one or more of a duration, an intensity, or a frequency of movement of the particular electronic device; and

inputting an input data set that includes the movement statistic to the trained machine-learning model.

24. The system of claim 19 , wherein the instructions further cause the one or more data processors to perform operations including:

determining that an alert condition is satisfied based on the result; and

as a result of determining that the alert condition is satisfied, transmitting the result to another electronic device.

25. The system of claim 19 , wherein the instructions further cause the one or more data processors to perform operations including:

accessing, by another electronic device, a user classification model based at least in part on health data for a plurality of users;

wherein determining that data is to be monitored at the particular electronic device for predictions of the event comprises:

classifying, by the other electronic device, a user into a user group using the user classification model; and

selecting, by the other electronic device, the trained machine-learning model based at least in part on the user group to which the user is classified.

26. The system of claim 25 , wherein the instructions further cause the one or more data processors to perform operations including:

executing, by the other electronic device, a function on at least a portion of the health data of the plurality of users to produce group expressions;

mapping, by the other electronic device, the group expressions to an N-dimensional space; and

classifying, by the other electronic device, a region bounded by the group expressions as a user group.

27. The system of claim 19 , wherein the trained machine-learning model is trained a machine-learning model with health record data of a user associated with the particular electronic device.