Methods and apparatus for anomaly detections

1. A system comprising:

a computing device configured to:

receive a plurality of features that include a cost feature;

generate a feature score for each of the plurality of features based on a log transformation of each of the plurality of features with respect to the cost feature, wherein at least a portion of the log transformations comprise a ratio between each respective feature of the plurality of features and the cost feature;

receive a request to update a value;

determine a machine learning model to apply to the value based on the received request, wherein the machine learning model is trained based on the generated feature scores;

generate an anomaly score based on application of the machine learning model to the value;

determine whether the request is an anomaly based on the anomaly score;

allow the update to the value if the request is determined not to be an anomaly; and

if the request is determined to be an anomaly:

deny the update to the value;

store anomaly data identifying the anomaly to a database; and

re-train the machine learning model based on the stored anomaly data.

2. The system of claim 1 , wherein determining whether the request is an anomaly based on the anomaly score comprises

determining whether the anomaly score is beyond a predetermined amount.

3. The system of claim 2 , wherein generating the anomaly score for the value is based on

A

i

⁡

(

x

i

)

=

Σ

i

∈

A

L

⁢

(

x

i

-

μ

i

)

2

σ

i

2

.

4. The system of claim 1 , wherein the plurality of features comprise at least one of a price feature, a binary feature, a categorical feature, and a hierarchical feature.

5. The system of claim 1 , wherein the machine learning model is an unsupervised machine learning model.

6. The system of claim 1 , wherein the value is at least one of a price and a cost of an item.

7. The system of claim 1 , wherein denying the update to the value if the request is determined to be the anomaly comprises:

generating a block update signal identifying that the update to the value is not to be allowed; and

transmitting the block update signal to a pricing system.

8. The system of claim 1 , wherein the computing device is configured to:

determine the request is the anomaly;

determine an impact score for the anomaly based on at least one of an estimated profit loss and a forgone revenue; and

determine a prioritization of the anomaly based on the determined impact score.

9. The system of claim 1 , wherein determining whether the request is an anomaly based on the anomaly score comprises comparing the anomaly score to a threshold.

10. A method comprising:

receiving a plurality of features that include a cost feature;

generating a feature score for each of the plurality of features based on a log transformation of each of the plurality of features with respect to the cost feature, wherein at least a portion of the log transformations comprise a ratio between each respective feature of the plurality of features and the cost feature;

receiving a request to update a value;

determining a machine learning model to apply to the value based on the received request, wherein the machine learning model is trained based on the generated feature scores;

generating an anomaly score based on application of the machine learning model to the value;

determining whether the request is an anomaly based on the anomaly score;

allowing the update to the value if the request is determined not to be an anomaly; and

if the request is determined to be an anomaly:

denying the update to the value;

storing anomaly data identifying the anomaly to a database; and

re-training the machine learning model based on the stored anomaly data.

11. The method of claim 10 wherein determining whether the request is an anomaly based on the anomaly score comprises:

generating an anomaly score for the value; and

determining whether the anomaly score is beyond a predetermined amount.

12. The method of claim 10 wherein the plurality of features comprising at least one of a price feature, a binary feature, a categorical feature, and a hierarchical feature.

13. The method of claim 10 wherein denying the update to the value if the request is determined to be the anomaly comprises:

generating a block update signal identifying that the update to the value is not to be allowed; and

transmitting the block update signal to a pricing system.

14. The method of claim 10 further comprising:

determining the request is the anomaly;

determining an impact score for the anomaly based on at least one of an estimated profit loss and a forgone revenue; and

determining a prioritization of the anomaly based on the determined impact score.

15. The method of claim 10 , wherein determining whether the request is an anomaly based on the anomaly score comprises comparing the anomaly score to a threshold.

16. A non-transitory computer readable medium having instructions stored thereon, wherein the instructions, when executed by at least one processor, cause a device to perform operations comprising:

receiving a plurality of features that include a cost feature;

generating a feature score for each of the plurality of features based on a log transformation of each of the plurality of features with respect to the cost feature, wherein at least a portion of the log transformations comprise a ratio between each respective feature of the plurality of features and the cost feature;

receiving a request to update a value;

determining a machine learning model to apply to the value based on the received request, wherein the machine learning model is trained based on the generated feature scores;

generating an anomaly score based on application of the machine learning model to the value;

determining whether the request is an anomaly based on the anomaly score;

allowing the update to the value if the request is determined not to be an anomaly; and

if the request is determined to be an anomaly:

denying the update to the value;

storing anomaly data identifying the anomaly to a database; and

re-training the machine learning model based on the stored anomaly data.

17. The non-transitory computer readable medium of claim 16 further comprising instructions stored thereon that, when executed by at least one processor, further cause the device to perform operations comprising

determining whether the anomaly score is beyond a predetermined amount.

18. The non-transitory computer readable medium of claim 16 further comprising instructions stored thereon that, when executed by at least one processor, further cause the device to perform operations comprising training the machine learning model based on at a plurality of features comprising at least one of a price feature, a binary feature, a categorical feature, and a hierarchical feature.

19. The non-transitory computer readable medium of claim 16 further comprising instructions stored thereon that, when executed by at least one processor, further cause the device to perform operations comprising:

determining the request is the anomaly;

determining an impact score for the anomaly based on at least one of an estimated profit loss and a forgone revenue; and

determining a prioritization of the anomaly based on the determined impact score.

20. The non-transitory computer readable medium of claim 16 , wherein determining whether the request is an anomaly based on the anomaly score comprises comparing the anomaly score to a threshold.