Techniques for data-driven correlation of metrics
Described herein are techniques for identifying highly relevant content for a user to view in the form of KPI cards and providing the relevant view to the user automatically or by suggestion. The KPIs of highest practical and statistical significance are provided when the user accesses the user interface. In some embodiments, when the user is viewing a KPI, other relevant KPIs may be provided for the user to view as suggestions. Further, in some embodiments, the user may be provided with the KPIs of significance based on anomaly detection, and the explanation for the anomaly as well as suggestions for remedying any issues may be provided to the user. The highly informational content can be surfaced through the use of the advanced machine learning algorithms described herein.
1. A computer-implemented method for identifying and displaying high informational content to a user, comprising:
for each metric indicator of a plurality of metric indicators of an enterprise:
training a model of the metric indicator using a machine learning process, the model having each attribute of a plurality of attributes of the metric indicator as independent variables of the model and configured to calculate derivatives for each of the plurality of attributes; and
selecting, based on the model, at least one attribute of the plurality of attributes having a greatest statistical change contribution using the model of the metric indicator, wherein the greatest statistical change contribution is based on the derivatives calculated by the model;
selecting a subset of the plurality of metric indicators from the plurality of metric indicators that have a greatest metric entropy change;
generating a deck of metric indicators based on the subset of the plurality of metric indicators;
generating a graphical display of each of the subset of the plurality of metric indicators, wherein each graphical display of the subset of the plurality of metric indicators comprises the at least one attribute of the plurality of attributes; and
providing the graphical display to a device of the user for display to the user.
2. The computer-implemented method of claim 1 , wherein selecting a subset of the plurality of metric indicators comprises:
obtaining a previous forecast of a current metric indicator;
determine a current value of the metric indicator; and
select the metric indicator when the previous forecast is different than the current value.
3. The computer-implemented method of claim 1 , the method further comprising:
generating an explanation of the greatest metric entropy change for the subset of the plurality of metric indicators by applying a current value of the metric indicator to the model, generating a partial derivative with respect to each of the plurality of attributes of the model, and selecting a subset of each of the plurality of attributes based on the generated partial derivative.
4. The computer-implemented method of claim 3 , the method further comprising:
running a what-if simulation of the model by changing values of the selected subset of each of the plurality of attributes to change the value of the metric indicator.
5. The method of claim 1 , the method further comprising:
recording actions of the user within a user interface;
generating user preference information of at least one metric indicator; and
wherein generating a deck of metric indicators comprises including the at least one metric indicator in the deck of metric indicators.
6. The method of claim 1 , wherein generating a graphical display of each of the subset of the plurality of metric indicators comprises:
creating a metric card comprising at least one visual depiction of the at least one attribute.
7. The method of claim 1 , further comprising:
identifying a second subset of the plurality of metric indicators to which the user has access; and
selecting the subset of the plurality of metric indicators from the second subset of the plurality of metric indicators.
8. The method of claim 1 , further comprising:
tracking usage behavior of a plurality of users using the user interface;
identifying trend data about each of the plurality of metric indicators based on the usage behavior of the plurality of users;
generating a graphical trend card comprising a visual depiction of the trend data; and
providing the graphical trend card to one or more users of the plurality of users in the user interface.
9. A system, comprising:
one or more processors; and
a memory having stored thereon instructions that, when executed by the one or more processors, causes the one or more processors to perform operations comprising:
for each metric indicator of a plurality of metric indicators of an enterprise:
training a model of the metric indicator using a machine learning process, the model having each attribute of a plurality of attributes of the metric indicator as independent variables of the model and configured to calculate derivatives for each of the plurality of attributes; and
selecting, based on the model, at least one attribute of the plurality of attributes having a greatest statistical change contribution using the model of the metric indicator, wherein the greatest statistical change contribution is based on the derivatives calculated by the model;
selecting a subset of the plurality of metric indicators from the plurality of metric indicators that have a greatest metric entropy change;
generating a deck of metric indicators based on the subset of the plurality of metric indicators;
generating a graphical display of each of the subset of the plurality of metric indicators, wherein each graphical display of the subset of the plurality of metric indicators comprises the at least one attribute of the plurality of attributes; and
providing the graphical display to a device of a user for display to the user.
10. The system of claim 9 , wherein the operations further comprise:
generating a dynamic dependency model using time-series data for each of the plurality of metric indicators.
11. The system of claim 10 , wherein the operations further comprise:
generating predictions of attribute values for each of the plurality of metric indicators using the dynamic dependency models.
12. The system of claim 11 , wherein the operations further comprise:
comparing the predictions of attribute values with actual attributes values to generate a statistical deviation using a divergence test for the actual attribute values.
13. The system of claim 12 , wherein selecting the subset of the plurality of metric indicators from the plurality of metric indicators that have a greatest metric entropy change comprises:
identifying the subset of the plurality of indicators as having a significant statistical deviation between the predictions of the attribute values and the actual attributes values.
14. The system of claim 13 , wherein selecting the subset of the plurality of metric indicators from the plurality of metric indicators further comprises:
selecting the subset of the plurality of metric indicators based on a user preference for metric indicators that have previously been viewed regularly.
15. A non-transitory computer-readable medium comprising instructions that, when executed by one or more processors, cause the one or more processors to perform operations comprising:
for each metric indicator of a plurality of metric indicators of an enterprise:
training a model of the metric indicator using a machine learning process, the model having each attribute of a plurality of attributes of the metric indicator as independent variables of the model and configured to calculate derivatives for each of the plurality of attributes; and
selecting, based on the model, at least one attribute of the plurality of attributes having a greatest statistical change contribution using the model of the metric indicator, wherein the greatest statistical change contribution is based on the derivatives calculated by the model;
selecting a subset of the plurality of metric indicators from the plurality of metric indicators that have a greatest metric entropy change;
generating a deck of metric indicators based on the subset of the plurality of metric indicators;
generating a graphical display of each of the subset of the plurality of metric indicators, wherein each graphical display of the subset of the plurality of metric indicators comprises the at least one attribute of the plurality of attributes; and
providing the graphical display to a device of a user for display to the user.
16. The non-transitory computer-readable medium of claim 15 , wherein the deck of metric indicators comprises a collection of graphical cards that provide data associated with the subset of metric indicators.
17. The non-transitory computer-readable medium of claim 16 , wherein graphical display of each of the subset of the plurality of metric indicators comprises a graphical dashboard displaying the collection of graphical cards.
18. The non-transitory computer-readable medium of claim 16 , wherein the collection of graphical cards display graphs of the data associated with the subset of metric indicators, wherein a type of the graphs is determined automatically by a user preference.
19. The non-transitory computer-readable medium of claim 16 , wherein the collection of graphical cards display actual attribute values and predicted attribute values.
20. The non-transitory computer-readable medium of claim 15 , wherein training data for the models of the metric indicators is selected from data stored in a data warehouse for the enterprise.