Machine learning platform for dynamic device and sensor quality evaluation
Aspects of the disclosure relate to computing platforms that utilize improved machine learning techniques for dynamic device quality evaluation. A computing platform may receive driving data from a mobile device. Using the driving data, the computing platform may compute a plurality of driving metrics, which may include: a geopoint expectation rate score, a trips per day rank score, a consecutive geopoint time difference score, a global positioning system (GPS) accuracy rating score, and a distance between consecutive trips score. By applying a machine learning model to the plurality of driving metrics, the computing platform may compute a device evaluation score, indicating a quality of the driving data received from the mobile device. Based on the device evaluation score, the computing platform may set flags, which may be accessible by a driver score generation platform, causing the driver score generation platform to perform an action with regard to the mobile device.
1 . A computing platform, comprising:
at least one processor;
a communication interface communicatively coupled to the at least one processor; and
memory storing instruction code that when executed by the at least one processor causes the computing platform to:
identify, for each of a plurality of consecutive driving trips specified within driving data, a time difference between an end point of a first driving trip and a starting point of a second driving trip;
determine a percentage of the time differences that exceed a predetermined period of time;
determine, based at least in part on a comparison of the percentage to a predetermined percentage value, a device evaluation score indicative of an accuracy of the driving data; and
selectively exclude the driving data from subsequent driving score computations when the device evaluation score is below a predetermined accuracy threshold, thereby reducing a likelihood of a driving score being computed based on inaccurate data.
2 . The computing platform of claim 1 , wherein the driving data specifies one or more driving metrics comprising: a geopoint expectation rate score, a trips per day rank score, a consecutive geopoint time difference score, a global positioning system (GPS) accuracy rating score, or a distance between consecutive trips score.
3 . The computing platform of claim 2 , wherein the instruction code for determining the device evaluation score causes the computing platform to:
apply the one or more driving metrics to a machine learning model trained to infer the device evaluation score.
4 . The computing platform of claim 1 , wherein the instruction code causes the computing platform to:
receive the driving data from a first computing device.
5 . The computing platform of claim 4 , wherein the instruction code causes the computing platform to:
identify a device model of the first computing device.
6 . The computing platform of claim 1 , wherein the instruction code causes the computing platform to:
communicate a message to an enterprise user device, wherein the message comprises an indication of the device evaluation score and one or more commands directing the enterprise user device to indicate the device evaluation score.
7 . The computing platform of claim 1 , wherein the instruction code for determining the device evaluation score causes the computing platform to:
determine the device evaluation score based at least in part on a median distance between respective end points and starting points of the consecutive driving trips.
8 . A non-transitory computer readable medium having stored thereon instruction code that, when executed by at least one processor of a computing platform, causes the computing platform to:
identify, for each of a plurality of consecutive driving trips specified within driving data, a time difference between an end point of a first driving trip and a starting point of a second driving trip;
determine a percentage of the time differences that exceed a predetermined period of time;
determine, based at least in part on a comparison of the percentage to a predetermined percentage value, a device evaluation score indicative of an accuracy of the driving data; and
selectively exclude the driving data from subsequent driving score computations when the device evaluation score is below a predetermined accuracy threshold, thereby reducing a likelihood of a driving score being computed based on inaccurate data.
9 . The non-transitory computer readable medium of claim 8 , wherein the driving data specifies one or more driving metrics comprising: a geopoint expectation rate score, a trips per day rank score, a consecutive geopoint time difference score, a global positioning system (GPS) accuracy rating score, or a distance between consecutive trips score.
10 . The non-transitory computer readable medium of claim 9 wherein the instruction code for determining the device evaluation score causes the computing platform to:
apply the one or more driving metrics to a machine learning model trained to infer the device evaluation score.
11 . The non-transitory computer readable medium of claim 8 , wherein the instruction code causes the computing platform to:
receive the driving data from a first computing device.
12 . The non-transitory computer readable medium of claim 11 , wherein the instruction code causes the computing platform to:
identify a device model of the first computing device.
13 . The non-transitory computer readable medium of claim 8 , wherein the instruction code causes the computing platform to:
communicate a message to an enterprise user device, wherein the message comprises an indication of the device evaluation score and one or more commands directing the enterprise user device to indicate the device evaluation score.
14 . The non-transitory computer readable medium of claim 8 , wherein the instruction code for determining the device evaluation score causes the computing platform to:
determine the device evaluation score based at least in part on a median distance between respective end points and starting points of the consecutive driving trips.
15 . A computer-implemented method comprising:
identifying, for each of a plurality of consecutive driving trips specified within driving data, a time difference between an end point of a first driving trip and a starting point of a second driving trip;
determining a percentage of the time differences that exceed a predetermined period of time;
determining, based at least in part on a comparison of the percentage to a predetermined percentage value, a device evaluation score indicative of an accuracy of the driving data; and
selectively excluding the driving data from subsequent driving score computations when the device evaluation score is below an predetermined accuracy threshold, thereby reducing a likelihood of a driving score being computed based on inaccurate data.
16 . The computer-implemented method of claim 15 , wherein the driving data specifies one or more driving metrics comprising: a geopoint expectation rate score, a trips per day rank score, a consecutive geopoint time difference score, a global positioning system (GPS) accuracy rating score, or a distance between consecutive trips score.
17 . The computer-implemented method of claim 16 , wherein determining the device evaluation score further comprises:
applying the one or more driving metrics to a machine learning model trained to infer the device evaluation score.
18 . The computer-implemented method of claim 15 , further comprising:
receiving the driving data from a first computing device.
19 . The computer-implemented method of claim 18 , further comprising:
identifying a device model of the first computing device.
20 . The computer-implemented method of claim 15 , further comprising:
communicating a message to an enterprise user device, wherein the message comprises an indication of the device evaluation score and one or more commands directing the enterprise user device to indicate the device evaluation score.