Utilizing predictive modeling to identify anomaly events
Approaches are provided for a prediction model determining a prediction indicative of the authorized user being associated with at least one anomaly event. In one example, a computer system may receive data from different sources. For example, the system may receive data associated with use of one or more monitored units of an automated storage and retrieval location. The system may also receive request data associated with a request for execution of the monitored controlled unit by an authorized user to a target user. The prediction model of the system may utilize the received data to determine the prediction that the authorized user is associated with at least one anomaly event associated with a diversion of a monitored unit away from the target user. The system may then provide the prediction for presentation.
1. A computer-implemented method, comprising:
receiving, by a computer system from one or more data sources, first data comprising a plurality of first data attributes associated with a monitored controlled unit of an automated storage and retrieval location, the automated storage and retrieval location comprising a storage unit in which is retained monitored controlled units, and the first data associated with a type of use involving the monitored controlled unit;
receiving, by the computer system from the one or more data sources, second data comprising a plurality of second data attributes corresponding to a request for the monitored controlled unit of the monitored controlled units retained in the automated storage and retrieval location, the request associated with an electronic record of a dependent user, and the second data associated with an order for the monitored controlled unit with respect to the dependent user;
receiving, by the computer system from the one or more data sources, third data comprising a plurality of third data attributes associated with an authorized user of a location that includes the automated storage and retrieval location;
determining, by a prediction model of the computer system, a prediction indicative of the authorized user being associated with at least one anomaly event, the prediction determined based at least in part on at least one of the first data, the second data, or the third data, and wherein the prediction model is trained based at least in part on features derived from at least one of the first data attributes, the second data attributes, or the third data attributes; and
providing, by the computer system, the prediction to a user device for presentation.
2. The computer-implemented method of claim 1 , wherein the authorized user is one of a plurality of authorized users of the location having a common role type, and wherein the method further comprises:
determining, by the computer system, a percentile score of the authorized user that relates the authorized user to other authorized users of the plurality of authorized users of the location having the common role type; and
training, by the computer system, the prediction model based at least in part on the percentile score.
3. The computer-implemented method of claim 2 , wherein the percentile score is associated with the type of use involving one or more monitored controlled units, the type of use including at least one of: (I) administration of the monitored controlled unit to the dependent user, (II) discarding of the monitored controlled unit, or (III) returning of the monitored controlled unit to the automated storage and retrieval location.
4. The computer-implemented method of claim 1 , wherein the prediction is presented via a graphical user interface of the user device that corresponds to a dashboard, the dashboard operable for ranking a level of the authorized user in comparison with levels of other authorized users of the location.
5. The computer-implemented method of claim 1 , wherein the prediction corresponds to a particular pattern of behavior indicative of the at least one anomaly event, the prediction model determining prediction based at least in part on the type of use involving the monitored controlled unit.
6. The computer-implemented method of claim 1 , further comprising:
receiving, by the computer system, a first number of data samples that are respectively determined to be associated with non-anomaly events;
receiving, by the computer system, a second number of data samples that are respectively determined to be associated with anomaly events, wherein the first number is greater than the second number;
selecting, by the computer system, a subset of the first number of data samples so that a difference between a third number of data samples of the subset and the second number of data samples is within a predetermined difference; and
training, by the computer system, the prediction model based at least in part on the subset of the first number of data samples and the second number of data samples.
7. The computer-implemented method of claim 1 , wherein the prediction model is trained based at least in part on a plurality of features, a feature of the plurality of features associated with at least one of: (I) a pattern of use by an authorized user involving one or more monitored controlled units, (II) a score associated with a condition of a dependent user before or after execution of a monitored controlled unit, or (III) a comparison between the pattern of use by an authorized user and a second pattern of use by another authorized user.
8. The computer-implemented method of claim 1 , wherein the plurality of first data attributes identify one or more of an identity of the dependent user, a quantity of the monitored controlled unit removed from the automated storage and retrieval location, a time at which the monitored controlled unit was removed from the automated storage and retrieval location, an identity of a user who removed the monitored controlled unit from the automated storage and retrieval location, or an identification of a use of the monitored controlled unit after removal.
9. The computer-implemented method of claim 1 , wherein the plurality of second data attributes identify one or more of a dosage associated with the request, a frequency associated with the request, a scale associated with the order, an identity of an authorized user associated with the request, an identity of the dependent user, a time at which the monitored controlled unit was administered to the dependent user, or a quantity of the monitored controlled unit administered to the dependent user.
10. The computer-implemented method of claim 1 , wherein the plurality of second data attributes identify an admission time associated with the dependent user, a discharge time associated with the dependent user, one or more tests performed on the dependent user, or results from one or more pain assessments performed on the dependent user.
11. One or more non-transitory computer-readable storage devices comprising computer-executable instructions that, when executed by one or more computer systems, cause the one or more computer systems to perform operations comprising:
receiving, from one or more data sources, first data comprising a plurality of first data attributes associated with a monitored controlled unit of an automated storage and retrieval location, the automated storage and retrieval location comprising a storage unit in which is retained monitored controlled units, and the first data associated with a type of use involving the monitored controlled unit;
receiving, from the one or more data sources, second data comprising a plurality of second data attributes corresponding to a request for the monitored controlled unit of the monitored controlled units retained in the automated storage and retrieval location, the request associated with an electronic record of a dependent user, and the second data associated with an order for the monitored controlled unit with respect to the dependent user;
receiving, from the one or more data sources, third data comprising a plurality of third data attributes associated with an authorized user of a location that includes the automated storage and retrieval location;
determining, by a prediction model, a prediction indicative of the authorized user being associated with at least one anomaly event, the prediction determined based at least in part on at least one of the first data, the second data, or the third data, and wherein the prediction model is trained based at least in part on features derived from at least one of the first data attributes, the second data attributes, or the third data attributes; and
providing the prediction to a user device for presentation.
12. The one or more non-transitory computer-readable storage devices of claim 11 , wherein the authorized user is one of a plurality of authorized users of the location having a common role type, and wherein the instructions further comprise performing operations comprising:
determining a percentile score of the authorized user that relates the authorized user to other authorized users of the plurality of authorized users of the location having the common role type; and
training the prediction model based at least in part on the percentile score.
13. The one or more non-transitory computer-readable storage devices of claim 11 , wherein receiving the first data comprises at least one of: (I) accessing the first data from an archive file generated by the automated storage and retrieval location, or (II) receiving a data stream from the automated storage and retrieval location that includes the first data.
14. The one or more non-transitory computer-readable storage devices of claim 11 , wherein the second data is received from an electronic data warehouse configured to retain the electronic record of the dependent user and electronic records of other users.
15. The one or more non-transitory computer-readable storage devices of claim 11 , wherein the prediction model is trained based at least in part on a plurality of features, a feature of the plurality of features associated with at least one of: (I) a pattern of use by an authorized user involving one or more monitored controlled units, (II) a score associated with a condition of a dependent user before or after execution of a monitored controlled unit, or (III) a comparison between the pattern of use by an authorized user and a second pattern of use by another authorized user.
16. A system, comprising:
a memory configured to store computer-executable instructions; and
one or more processors configured to access the memory and execute the computer-executable instructions to at least:
receive, from one or more data sources, first data comprising a plurality of first data attributes associated with a monitored controlled unit of an automated storage and retrieval location, the automated storage and retrieval location comprising a storage unit in which is retained monitored controlled units, and the first data associated with a type of use involving the monitored controlled unit;
receive, from one or more data sources, second data comprising a plurality of second data attributes corresponding to a request for the monitored controlled unit of the monitored controlled units retained in the automated storage and retrieval location, the request associated with an electronic record of a dependent user, and the second data associated with an order for the monitored controlled unit with respect to the dependent user;
receive, from one or more data sources, third data comprising a plurality of third data attributes associated with an authorized user of a location that includes the automated storage and retrieval location;
determine, by a prediction model, a prediction indicative of the authorized user being associated with at least one anomaly event, the prediction determined based at least in part on at least one of the first data, the second data, or the third data, and wherein the prediction model is trained based at least in part on features derived from at least one of the first data attributes, the second data attributes, or the third data attributes; and
provide the prediction to a user device for presentation.
17. The system of claim 16 , wherein the automated storage and retrieval location is a computer-controlled controlled storage and retrieval location.
18. The system of claim 16 , wherein the authorized user was located at the location including the automated storage and retrieval location at a first time when the monitored controlled unit was removed from the automated storage and retrieval location, and wherein the one or more data sources includes at least one of a resource application, a record application, or a location system.
19. The system of claim 16 , wherein the plurality of first data attributes are further associated with a plurality of automated storage and retrieval locations, and wherein the plurality of second data attributes further correspond to a plurality of requests for a plurality of monitored controlled units.
20. The system of claim 16 , wherein the prediction is presented via a graphical user interface of the user device that corresponds to a dashboard, the dashboard operable for ranking a level of the authorized user in comparison with levels of other authorized users of the location.