ENSEMBLE MODELS FOR ANOMALY DETECTION
The subject technology detects anomalies in media campaign configuration settings. The anomaly detection system may leverage one or more deep learning models to detect anomalies and identify particular configuration settings that contribute to the detected anomalies. In various embodiments, two or more of the deep learning models may be combined into an ensemble model that boosts the accuracy of anomaly predictions made by the anomaly detection system. The anomaly detection system may review the configuration settings of media campaigns during the configuration process and before the media campaigns run on a publication system in order to reduce the amount of unsuccessful campaigns and minimize the amount of wasted resources spent on running campaigns that have a low likelihood of achieving user defined goals.
1 . An anomaly detection system comprising:
one or more processors; and
a memory storing instructions that, when executed by at least one processor in the one or more processors, cause the at least one processor to perform at least the following operations:
encode one or more configuration settings for a campaign in a composite vector, the composite vector being representative of at least a portion of the configuration settings;
generate a composite vector for multiple campaigns to create a training sample including a set of composite vectors;
train multiple anomaly detection machine learning models using the training sample; and
determine, using an ensemble model, an anomaly occurrence probability for one or more configuration settings of a target campaign, the ensemble model including two or more of the trained anomaly detection machine learning models.
2 . The system of claim 1 , wherein the at least one or more processors are further configured to determine a composite vector for the target campaign, the composite vector being begin representative of at least a portion of the configuration settings of the target campaign.
3 . The system of claim 1 , wherein the at least one processor is further configured to use each of the two or more trained models to determine an anomaly occurrence probability for one or more of the configuration settings of the target campaign; and
use a weighted voting process to aggregate the anomaly occurrence probabilities determined by the two or more trained models to generate a composite anomaly occurrence probability for the one or more configuration settings of the target campaign.
4 . The system of claim 2 , wherein the composite vector for the target campaign includes normalized features determined for one or more numeric configuration settings and encoded features determined from one or more categorical configuration settings.
5 . The system of claim 1 , wherein the two or more trained models determine the anomaly occurrence probability by determining a numerical probability that the target campaign includes an anomaly and comparing the numerical probability to an anomaly threshold.
6 . The system of claim 3 , wherein the weighted voting process multiples the anomaly occurrence probability determined by each of the trained models by a voting weight determined based on an anomaly detection accuracy of each of the trained models measured for a test sample of completed campaigns.
7 . The system of claim 1 , wherein the at least one processor is further configured to identify an anomaly based on the anomaly occurrence probability for one or more configuration settings of the target campaign.
8 . The system of claim 7 , wherein the at least one processor is further configured to identify the one or more configuration settings that contribute to the anomaly using an output of one or more of the trained models, wherein the output used to identify the one or more configuration settings that contribute to the anomaly includes a joint probability distribution determined from an output of one or more layers of at least one of the trained models.
9 . The system of claim 7 , wherein the at least one processor is further configured to display an error message that includes the identified one or more configuration settings that contribute to the anomaly.
10 . A method of anomaly detection in media campaign configuration settings, the method comprising:
encoding one or more configuration settings for a campaign in a composite vector, the composite vector being representative of at least a portion of the configuration settings;
generating a composite vector for multiple campaigns to create a training sample including a set of composite vectors;
training multiple anomaly detection machine learning models using the training sample; and
determining, using an ensemble model, an anomaly occurrence probability for one or more configuration settings of a target campaign, the ensemble model including two or more of the trained anomaly detection machine learning models.
11 . The method of claim 10 , further comprising determining a composite vector for the target campaign, the composite vector being begin representative of at least a portion of the configuration settings of the target campaign.
12 . The method of claim 10 , further comprising using each of the two or more trained models to determine an anomaly occurrence probability for one or more of the configuration settings of the target campaign; and
using a weighted voting process to aggregate the anomaly occurrence probabilities determined by the two or more trained models to generate a composite anomaly occurrence probability for the one or more configuration settings of the target campaign.
13 . The method of claim 10 , wherein the composite vector for the target campaign includes normalized features determined for one or more numeric configuration settings and encoded features determined from one or more categorical configuration settings.
14 . The method of claim 10 , further comprising determining the anomaly occurrence probability by determining a numerical probability that the target campaign includes an anomaly and comparing the numerical probability to an anomaly threshold.
15 . The method of claim 12 , further comprising performing the weighted voting process by multiplying the anomaly occurrence probability determined by each of the trained models by a voting weight determined based on an anomaly detection accuracy of each of the trained models measured for a test sample of completed campaigns.
16 . The method of claim 10 , further comprising identifying an anomaly based on the anomaly occurrence probability for one or more configuration settings of the target campaign.
17 . The method of claim 16 , further comprising identifying the one or more configuration settings that contribute to the anomaly using an output of one or more of the trained models, wherein the output used to identify the one or more configuration settings that contribute to the anomaly includes a joint probability distribution determined from an output of one or more layers of at least one of the trained models.
18 . The method of claim 16 , further comprising displaying an error message that includes the identified one or more configuration settings that contribute to the anomaly.
19 . A method of anomaly detection in media campaign configuration settings, the method comprising:
encoding one or more configuration settings for a campaign in a composite vector, the composite vector being representative of at least a portion of the configuration settings;
generating a composite vector for multiple campaigns to create a training sample including a set of composite vectors;
training an anomaly detection machine learning model using the training sample;
determining, using the trained anomaly detection machine learning model, a joint probability distribution for multiple sets of at least two configuration settings of a target campaign; and
determining a conditional probability for each set of at least two configuration settings based on the joint probability distributions.
20 . The method of claim 19 further comprising identifying a set of at least two or more configuration settings that contain an anomaly based on the conditional probabilities.