Systems and methods for match evaluation based on change in telematics inferences via a telematics marketplace
A computer implemented including: collecting, via one or more sensors, a first set of operator data associated with a first group of vehicle operators during a first time period; determining, for each vehicle operator of the first group of vehicle operators and using one or more trained machine learning models based at least in part upon the first set of operator data, a first set of telematics inferences, the one or more trained machine learning models being trained using training data sets comprising sensor data associated with a second group of vehicle operators to predict telematics inferences; collecting, via the one or more sensors, a second set of operator data associated with the first group of vehicle operators during a second time period; determining, for each vehicle operator of the first group of vehicle operators and based on the second set of operator data, a second set of telematics inferences; determining, for each vehicle operator of the first group of vehicle operators, one or more match evaluations based at least in part upon the first set of telematics inferences and the second set of telematics inferences; and modifying one or more weights of the one or more trained machine learning models based at least in part upon the one or more match evaluations. Other embodiments are described.
1. A computer-implemented method comprising:
collecting, via one or more sensors, a first set of operator data associated with a first group of vehicle operators during a first time period;
determining, for each vehicle operator of the first group of vehicle operators and using one or more trained machine learning models based at least in part upon the first set of operator data, a first set of telematics inferences, the one or more trained machine learning models trained using training data sets comprising sensor data associated with a second group of vehicle operators to predict telematics inferences, and the one or more trained machine learning models comprising one or more of a reinforcement learning model or a convolutional neural network;
collecting, via the one or more sensors, a second set of operator data associated with the first group of vehicle operators during a second time period;
determining and updating, for each vehicle operator of the first group of vehicle operators and based on the second set of operator data, a second set of telematics inferences;
determining and updating, for each vehicle operator of the first group of vehicle operators, one or more match evaluations based at least in part upon one or more differences between the first set of telematics inferences and the second set of telematics inferences; and
modifying one or more weights of the one or more trained machine learning models based at least in part upon the one or more match evaluations associated with the differences between the first set of telematics inferences and the second set of telematics inferences.
2. The computer-implemented method of claim 1 , wherein the first set of telematics inferences and the second set of telematics inferences comprise, respectively, one or more of:
a predicted profitability and an actual profitability;
a predicted period of retention and an actual period of retention;
a predicted revenue and an actual revenue;
predicted costs and actual costs; or
one or more predicted behavioral scores and one or more actual behavioral scores.
3. The computer-implemented method of claim 1 , wherein determining the one or more match evaluations comprises determining whether one or more actual metrics of the second set of telematics inferences exceed, meet, or fall below one or more predicted metrics of the first set of telematics inferences.
4. The computer-implemented method of claim 1 , wherein:
the first time period occurs prior to a user acquisition by one or more marketplace participants; and
the second time period occurs after the user acquisition by the one or more marketplace participants.
5. The computer-implemented method of claim 1 , wherein the one or more sensors are controlled by one or more mobile applications used by the first group of vehicle operators.
6. The computer-implemented method of claim 5 , wherein the one or more mobile applications comprise a system software application, an entertainment software application, a gaming software application, a navigation software application, or an environment software application.
7. A system comprising one or more processors and one or more non-transitory computer-readable media storing computing instructions that, when executed by the one or more processors, cause the one or more processors to perform operations comprising:
collecting, via one or more sensors, a first set of operator data associated with a first group of vehicle operators during a first time period;
determining, for each vehicle operator of the first group of vehicle operators and using one or more trained machine learning models based at least in part upon the first set of operator data, a first set of telematics inferences, the one or more trained machine learning models trained using training data sets comprising sensor data associated with a second group of vehicle operators to predict telematics inferences, and the one or more trained machine learning models comprising one or more of a reinforcement learning model or a convolutional neural network;
collecting, via the one or more sensors, a second set of operator data associated with the first group of vehicle operators during a second time period;
determining and updating, for each vehicle operator of the first group of vehicle operators and based on the second set of operator data, a second set of telematics inferences;
determining and updating, for each vehicle operator of the first group of vehicle operators, one or more match evaluations based at least in part upon one or more differences between the first set of telematics inferences and the second set of telematics inferences; and
modifying one or more weights of the one or more trained machine learning models based at least in part upon the one or more match evaluations associated with the differences between the first set of telematics inferences and the second set of telematics inferences.
8. The system of claim 7 , wherein the first set of telematics inferences and the second set of telematics inferences comprise, respectively, one or more of:
a predicted profitability and an actual profitability;
a predicted period of retention and an actual period of retention;
a predicted revenue and an actual revenue;
predicted costs and actual costs; or
one or more predicted behavioral scores and one or more actual behavioral scores.
9. The system of claim 7 , wherein determining the one or more match evaluations comprises determining whether one or more actual metrics of the second set of telematics inferences exceed, meet, or fall below one or more predicted metrics of the first set of telematics inferences.
10. The system of claim 7 , wherein:
the first time period occurs prior to a user acquisition by one or more marketplace participants; and
the second time period occurs after the user acquisition by the one or more marketplace participants.
11. The system of claim 7 , wherein the one or more sensors are controlled by one or more mobile applications used by the first group of vehicle operators.
12. The system of claim 11 , wherein the one or more mobile applications comprise a system software application, an entertainment software application, a gaming software application, a navigation software application, or an environment software application.
13. One or more non-transitory computer-readable media storing computing instructions that, when executed by one or more processors, cause the one or more processors to perform operations comprising:
collecting, via one or more sensors, a first set of operator data associated with a first group of vehicle operators during a first time period;
determining, for each vehicle operator of the first group of vehicle operators and using one or more trained machine learning models based at least in part upon the first set of operator data, a first set of telematics inferences, the one or more trained machine learning models trained using training data sets comprising sensor data associated with a second group of vehicle operators to predict telematics inferences, and the one or more trained machine learning models comprising one or more of a reinforcement learning model or a convolutional neural network;
collecting, via the one or more sensors, a second set of operator data associated with the first group of vehicle operators during a second time period;
determining and updating, for each vehicle operator of the first group of vehicle operators and based on the second set of operator data, a second set of telematics inferences;
determining and updating, for each vehicle operator of the first group of vehicle operators, one or more match evaluations based at least in part upon one or more differences between the first set of telematics inferences and the second set of telematics inferences; and
modifying one or more weights of the one or more trained machine learning models based at least in part upon the one or more match evaluations associated with the differences between the first set of telematics inferences and the second set of telematics inferences.
14. The one or more non-transitory computer-readable media of claim 13 , wherein the first set of telematics inferences and the second set of telematics inferences comprise, respectively, one or more of:
a predicted profitability and an actual profitability;
a predicted period of retention and an actual period of retention;
a predicted revenue and an actual revenue;
predicted costs and actual costs; or
one or more predicted behavioral scores and one or more actual behavioral scores.
15. The one or more non-transitory computer-readable media of claim 13 , wherein determining the one or more match evaluations comprises determining whether one or more actual metrics of the second set of telematics inferences exceed, meet, or fall below one or more predicted metrics of the first set of telematics inferences.
16. The one or more non-transitory computer-readable media of claim 13 , wherein:
the first time period occurs prior to a user acquisition by one or more marketplace participants; and
the second time period occurs after the user acquisition by the one or more marketplace participants.
17. The one or more non-transitory computer-readable media of claim 13 , wherein the one or more sensors are controlled by one or more mobile applications used by the first group of vehicle operators.