Abstract: A computer-implemented system to detect vulnerabilities in artificial intelligence (AI) models, the system comprising a first AI model for calculating a first score for a first transaction based on one or more features extracted from the first transaction and transaction history associated with the first transaction, the first transaction being tagged as potentially adversarial, in response to determining that the first score is in an improbable range based on comparing first attributes associated with the first transaction with second attributes associated with at least a second transaction, the comparison indicating the first transaction has a low likelihood of occurrence; and a second AI model for identifying adversarial transactions, in response to determining that number of plurality of example transactions scored by the first model is sufficient to train the second AI model.

Abstract: A method includes generating a plurality of binary feature maps containing a set of feature map values including a first binary value and/or a second binary value, by at least converting each input value of a set of input values of a plurality of input feature vectors to the first binary value when the corresponding input value is the zero value or the second binary value when the corresponding input value is the non-zero value. The method includes segmenting the plurality of binary feature maps into a plurality of segments representing behavior profiles. Each segment includes at least one subsegment in which the set of feature map values is the same for all binary feature maps in the at least one subsegment. The method includes predicting, based on a segment of the plurality of segments, a specific outcome. Related methods and articles of manufacture are also disclosed.

Abstract: Systems for improving security of a computer-implemented artificial intelligence by monitoring one or more transactions received by the machine learning decision model; receiving a first score generated by the machine learning decision model in association with a first transaction; identifying the first transaction as belonging to a first class, in response to the first score being lower than a certain score threshold and the first transaction having a low occurrence likelihood; receiving a second score in association with the first transaction based on one or more adversarial latent features associated with the first transaction as detectable by an adversary detection model; and determining at least one adversarial latent transaction feature being exploited by the first transaction, in response to determining that the second score falls above the certain score threshold.

Abstract: A computer-implemented method for generating a classifier, comprising: assigning a plurality of hierarchies of tags to a collection of training examples, wherein a higher level tag of the plurality of hierarchies of tags comprises a set of lower level tags; associating, in the classifier, a plurality of latent features with each of the plurality of hierarchies of tags, respectively; constructing a plurality of loss functions, wherein each loss function is associated with each level of the plurality of hierarchies of tags and associated latent features of the classifier, wherein the loss function aggregates a plurality of binary cross entropy for each member of a level of tags and associated latent features; and training the classifier by minimizing the loss functions for each level of the plurality of hierarchies of tags and associated latent features of the classifier.

Abstract: A system and method for analyzing coverage, bias and model explanations in large dimensional modeling data includes discretizing three or more variables of a dataset to generate a discretized phase space represented as a grid of a plurality of cells, the dataset comprising a plurality of records, each record of the plurality of records having a value and a unique identifier (ID). A grid transformation is applied to each record in the dataset to assign each record to a cell of the plurality of cells of the grid according to the grid transformation. A grid index is generated to reference each cell using a discretized feature vector. A grid storage for storing the records assigned to each cell of the grid is then created. The grid storage using the ID of each record as a reference to each record and the discretized feature vector as a key to each cell.

Abstract: A system and method for analyzing coverage, bias and model explanations in large dimensional modeling data includes discretizing three or more variables of a dataset to generate a discretized phase space represented as a grid of a plurality of cells, the dataset comprising a plurality of records, each record of the plurality of records having a value and a unique identifier (ID). A grid transformation is applied to each record in the dataset to assign each record to a cell of the plurality of cells of the grid according to the grid transformation. A grid index is generated to reference each cell using a discretized feature vector. A grid storage for storing the records assigned to each cell of the grid is then created. The grid storage using the ID of each record as a reference to each record and the discretized feature vector as a key to each cell.

Abstract: A method is provided for multivariate counterfactual diffusion in desensitizing behavior latent features learned on limited data. The method includes generating a plurality of synthetic vectors for each input vector of a plurality of input vectors used to train a first machine learning model, where the plurality of synthetic vectors represent potential counterfactuals associated with the corresponding input vector. The method also includes filtering the plurality of synthetic vectors to identify counterfactual synthetic vectors. The method further includes predicting, by a second machine learning model trained based on the plurality of input vectors and the filtered plurality of counterfactual synthetic vectors, a classification of at least one input vector of the plurality of input vectors. Related methods and articles of manufacture are also disclosed.

Abstract: A method includes determining, by a trained machine learning model, a score based at least on one or more latent features. The method also includes monitoring the determining of the score by the trained machine learning model. The monitoring includes determining one or more production statistics associated with the one or more latent features, derived variables and input data elements, and accessing one or more reference assets persisted on a model governance blockchain. The one or more reference assets includes one or more reference statistics and a threshold indicating a deviation between the one or more production statistics and the one or more reference statistics. The method also includes generating an alert based on the one or more production statistics associated with the one or more latent features meeting the threshold. Related methods and articles of manufacture are also disclosed.

Abstract: A method includes generating a plurality of binary feature maps containing a set of feature map values including a first binary value and/or a second binary value, by at least converting each input value of a set of input values of a plurality of input feature vectors to the first binary value when the corresponding input value is the zero value or the second binary value when the corresponding input value is the non-zero value. The method includes segmenting the plurality of binary feature maps into a plurality of segments representing behavior profiles. Each segment includes at least one subsegment in which the set of feature map values is the same for all binary feature maps in the at least one subsegment. The method includes predicting, based on a segment of the plurality of segments, a specific outcome. Related methods and articles of manufacture are also disclosed.

Abstract: A method is provided for a first to saturate single modal latent feature activation network. The method includes training, based on a plurality of training examples including a plurality of input features, a first machine learning model including a hidden node. The method includes determining a plurality of subsets of the plurality of input features including a minimum combination of the plurality of input features first to cause saturation of the hidden node. The method includes determining a hidden node ordered saturation list including a subset of the plurality of subsets. The method includes generating a sparsely trained machine learning model to determine an output for a training example of the plurality of training examples based on at least one input feature of the subset included in the hidden node ordered saturation list corresponding to the hidden node. Related methods and articles of manufacture are also disclosed.

Abstract: A method may include generating synthetic data based on input data and training a machine learning model based on the synthetic data. The synthetic data may be generated by determining a plurality of data points representing an archetype probability distribution of a plurality of archetypes, clustering the plurality of data points into one or more clusters associated with transactional behavior patterns, generating a threshold metric representing a peak distribution density of the plurality of data points associated with a corresponding cluster, removing, from the plurality of data points, one or more non-representative data points to define a reduced set of the plurality of data points, generating an updated archetype probability distribution based at least on the reduced set of the plurality of data points, and generating representative transaction data based on the updated archetype probability distribution and threshold metric. Related methods and articles of manufacture are al so disclosed.

Abstract: A diagnostic system for model governance is presented. The diagnostic system includes an auto-encoder to monitor model suitability for both supervised and unsupervised models. When applied to unsupervised models, the diagnostic system can provide a reliable indication on model degradation and recommendation on model rebuild. When applied to supervised models, the diagnostic system can determine the most appropriate model for the client based on a reconstruction error of a trained auto-encoder for each associated model. An auto-encoder can determine outliers among subpopulations of consumers, as well as support model go-live inspections.

Abstract: Systems and methods for providing insights about a machine learning model are provided. The method includes, using training data to train the machine learning model to learn patterns to determine whether data associated with an event provides an indication that the event belongs to a certain class from among a plurality of classes, evaluating one or more features of the machine learning model to produce a data set pairing observed scores S and a set of predictive input variables Vi, and constructing at least one data-driven estimator based on an explanatory statistic, the estimator being represented in a computationally efficient form and packaged with the machine learning model and utilized to provide a definition of explainability for a score generated by the machine learning model.

Abstract: Systems and methods for generating concise explanations of scored observations that strike good, and computationally efficient, trade-offs between rank-ordering performance and explainability of scored observations are disclosed. The systems and methods described herein for explaining scored observations are based on a framework of partial dependence functions (PDFs), multi-layered neural networks (MNNs), and Latent Explanations Neural Network Scoring (LENNS).

Abstract: Systems for improving security of a computer-implemented artificial intelligence by monitoring one or more transactions received by the machine learning decision model; receiving a first score generated by the machine learning decision model in association with a first transaction; identifying the first transaction as belonging to a first class, in response to the first score being lower than a certain score threshold and the first transaction having a low occurrence likelihood; receiving a second score in association with the first transaction based on one or more adversarial latent features associated with the first transaction as detectable by an adversary detection model; and determining at least one adversarial latent transaction feature being exploited by the first transaction, in response to determining that the second score falls above the certain score threshold.

Abstract: Systems and methods for providing insights about a machine learning model are provided. The method includes, using training data to train the machine learning model to learn patterns to determine whether data associated with an event provides an indication that the event belongs to a certain class from among a plurality of classes, evaluating one or more features of the machine learning model to produce a data set pairing observed scores S and a set of predictive input variables Vi, and constructing at least one data-driven estimator based on an explanatory statistic, the estimator being represented in a computationally efficient form and packaged with the machine learning model and utilized to provide a definition of explainability for a score generated by the machine learning model.

Abstract: A diagnostic system for model governance is presented. The diagnostic system includes an auto-encoder to monitor model suitability for both supervised and unsupervised models. When applied to unsupervised models, the diagnostic system can provide a reliable indication on model degradation and recommendation on model rebuild. When applied to supervised models, the diagnostic system can determine the most appropriate model for the client based on a reconstruction error of a trained auto-encoder for each associated model. An auto-encoder can determine outliers among subpopulations of consumers, as well as support model go-live inspections.

Abstract: Systems, methods and computer program products for improving security of artificial intelligence systems. The system comprising processors for monitoring one or more transactions received by a machine learning decision model to determine a first score associated with a first transaction. The first transaction may be identified as likely adversarial, in response to the first score being lower than a certain score threshold and the first transaction having a low occurrence likelihood. A second score may be generated in association with the first transaction based on one or more adversarial latent features associated with the first transaction. At least one adversarial latent feature may be detected as being exploited by the first transaction, in response to determining that the second score falls above the certain score threshold. Accordingly, an abnormal volume of activations of adversarial latent features spanning across a plurality of transactions scored may be detected and blocked.

Abstract: A method, a system, and a computer program product for detecting a diverse set of rare behavior. A time-series data representing one or more actions executed by an entity is received from a plurality of time-series data sources and is processed. A data structure corresponding to the entity, identifying the entity, and including one or more representations of processed time-series data identifying the actions is generated. A current action executed by the entity is detected. Current time-series data corresponding to the current action is received and associated with the data structure. First features are extracted from the generated data structure based on current time-series data and compared to second features extracted for at least another entity to determine difference parameters between first and second features. One or more models are trained using difference parameters, and a score for each action executed by the entity is determined.

Abstract: Systems and methods for generating concise explanations of scored observations that strike good, and computationally efficient, trade-offs between rank-ordering performance and explainability of scored observations are disclosed. The systems and methods described herein for explaining scored observations are based on a framework of partial dependence functions (PDFs), multi-layered neural networks (MNNs), and Latent Explanations Neural Network Scoring (LENNS).