Abstract: A first set of variables is introduced into a data transaction scoring system. The first set of variables having a maturity level less than a second set of variables previously matured on the data transaction scoring system. The maturity level corresponding to an amount of exposure to the data transaction scoring system. The amount of exposure affecting a degree of precision for the data transaction scoring system. The first set of variables are introduced by expanding a data transaction profile for the data transaction scoring system including second set of variables to further include the first set of variables. Initial values are assigned for the first set of variables based on a statistical model. After such an assignment, scoring of data transactions using the expanded data transaction profile prior to maturing the data transaction system using the first set of variables can be initiated. Related apparatus, systems, techniques, and articles are also described.

Abstract: Methods and apparatus, including computer program products, for cardholder localization based on transaction data. In general, a determination of one or more possible base locations of a cardholder of a payment card may be initiated. Transactions of the cardholder may be associated with merchant customer base location profiles, where each of the merchant customer base location profiles models a distribution of base locations of customers with which a merchant has had a transaction. The one or more possible cardholder base locations may be derived from the merchant customer base location profiles. The one or more possible cardholder base locations may be used to determine a home or office location of a cardholder where such information is not otherwise available, to determine whether a cardholder has moved from a stated address, to determine if a cardholder has multiple residences, to detect fraud, and the like.

Abstract: A request to generate a consolidation risk score that characterizes a likelihood of a change in a level of creditworthiness of an individual following a consolidation of debt of the individual using a secured line of credit is received. Thereafter, future credit balance increases are estimated for the individual using a predictive model trained using historical creditworthiness data of a plurality of consolidators. These estimated future balance increases are then associated with a consolidation risk score so that such score can be provided. Related apparatus, systems, techniques, and articles are also described.

Abstract: A method for dynamically updating a model is described. The method includes accessing a model that specifies expected characteristics for a transaction. The model includes variables associated with fraud. The method also includes receiving at least one value for each of the variables while monitoring transactions, and updating a distribution of values for each variable based on the received value. The received value is compared with the updated distribution to determine a deviation from a threshold value associated with a percentile of the updated distribution that is indicative of fraud.

Abstract: The invention, referred to herein as PeaCoCk, uses a unique blend of technologies from statistics, information theory, and graph theory to quantify and discover patterns in relationships between entities, such as products and customers, as evidenced by purchase behavior. In contrast to traditional purchase-frequency based market basket analysis techniques, such as association rules which mostly generate obvious and spurious associations, PeaCoCk employs information-theoretic notions of consistency and similarity, which allows robust statistical analysis of the true, statistically significant, and logical associations between products. Therefore, PeaCoCk lends itself to reliable, robust predictive analytics based on purchase-behavior.

Abstract: A fraud detection and protection method and system are disclosed. The method and system utilize a fraud detection and protection server to monitor online commercial transactions between a webserver and a client computer, and generate a risk assessment of a user associated with the client computer. The system and method further utilize a number of beacon servers geographically dispersed in an area. Each beacon server is configured to receive packet header information associated with the online commercial transactions, analyze the packet header information for authenticating information, and send the authenticating information to the fraud detection and protection server for the risk assessment.

Abstract: An exemplary Web-based score explanation service typically requires only the credit bureau identifier, credit score, and up to four reason codes as input.

Abstract: A system and method for protecting identity fraud are disclosed. A system includes a detection subsystem to identify applications and/or accounts at risk of identity fraud, and a disposition subsystem to process data provided by the detection system and to determine whether identity fraud exists in the applications and/or accounts. According to an implementation, one or more neural network models are defined, each neural network model being configured to handle a class of cases related to the subject and a specific data configuration describing a case of the class. The one or more neural network models are run to generate data requests about the subject's identity, and the data requests are passed to a detection system that monitor transactions associated with the subject. Additional data associated with the transactions is requested until a threshold certainty is achieved or until available data or models are exhausted.

Abstract: A method and system for predicting impact of future actions on subsequent performance involves developing a prediction model that predicts a statistical interaction of performance expectation with likely behavior. In one embodiment, sensitivity to new, post-scoring date credit behaviors in the analytic solution greatly improves snapshot score predictions. The modeling approach involves multiple snapshots: predictive and performance snapshots, plus an intermediate snapshot shortly after the predictive snapshot to quantify interim behavior. Predictive interaction variables are calculated on the predictive data using simulated profiles before and after an action.

Abstract: A network assurance analytics (NAA) system and method is disclosed. The NAA can be part of a risk analytic for telecom (RAFT) program. The NAA system is configured to monitor telecommunications networks, detect errors or fraud in those telecommunications networks, and provide solutions to resolve the errors or reduce the fraud. Traffic of a telecommunications network is electronically monitored for at least one pattern that is indicative of a telecommunications anomaly. Based on a set of a set of telecommunications profiles stored in a database, a model score representing a value of the telecommunications anomaly is generated. A solution for the telecommunications network to reduce the model score associated with the telecommunications anomaly is then generated for execution on the telecommunications network.

Abstract: The invention provides a purchase sequence browser (PuSB), i.e. a graphical user interface (GUI) that facilitates insight discovery and exploration of product affinities across time, generated by a purchase sequence analysis of a retailer's transaction data. A purchase sequence browser allows the user to browse the most significant product phrases discovered by an exhaustive search of the product affinities across time; explore the retail grammar to create both forward and backward phrase trees or alternate purchase paths starting from, or ending in, a product; generate consistent purchase sequences given some constraints on the products and their order; and profile the value of a product across time with regard to other products fixed in time.

Abstract: A method and system for estimating potential future outcomes resulting from decision alternatives is presented to enable lenders to make lending related decisions. The estimation is based on a propensity score variable that encompasses an effect of multiple covariates associated with one or more individuals for whom the estimation is being performed. For consistency with empirical testing, the estimation approach assumes conditions of unconfoundedness and localized common support. According to the unconfoundedness assumption, for a given variable, the potential outcomes are conditionally independent of the decision alternatives. According to the localized common support assumption, an overlap is ensured between individual accounts that are categorized together as potentially having the same future outcome. The outcomes and an effect (e.g. comparison) of the outcomes may be displayed graphically.

Abstract: Disclosed is a data collection and analysis system that is capable of extracting data from various disparate sources, i.e., contact channels, storing the data and analyzing the data to show trends in the business operation. The data is stored in a data model that uses a star schema approach to providing a unified data source. Analyzed data can be made available to users on a nearly real time basis that allows the users to view trends in business operation and plan accordingly.

Abstract: A system and method for managing mass compromise of financial transaction devices is disclosed. A method includes maintaining a summary of a transaction history for a financial transaction device, and forming a device history profile based on the transaction history, the device history profile including predictive variables indicative of fraud associated with the financial transaction device. A method further includes generating a fraud score based on the predictive variables, the fraud score representing a likelihood that the financial transaction device is compromised will be used fraudulently.

Abstract: The invention comprises an article and method for transforming a complex or large decision tree having multiple variables; multiple values for each variable; and, multiple outcomes for each combination of variables and their associated values, into a compact, efficient graphical representation to provided enhanced ease of use and interaction by a human user. More particularly, the invention comprises a computationally efficient method for transforming an input decision tree into an optimal compact representation by computing a particular ordering of variables in the decision tree that first leads to a Directed Acyclic Graph, or “DAG,” with a minimum number of nodes. The method then converts the DAG into an exception-based DAG, or “EDAG,” with exactly one exception, having an optimal, minimum number of nodes with increased comprehensibility for a user.

Abstract: The invention provides an overall architecture for optimal strategy design using both historical data and human expertise. The presently preferred architecture supports the tasks of strategy design and strategy analysis, and provides extensions to influence diagrams, translation of an influence diagram as a nonlinear optimization problem, and use of induction after optimization to derive decision rules.

Abstract: Data spiders, provide an automated system that can take a file or file store of historic transaction data and create the best set of variables from that data, where “best” means highly predictive. Genetic algorithms are used to parameterized transactions to form groups, which are subjected naïve Bayes score ranking. Variable groups are generated and ranked accord to the score. Data spiders span the full information available, are uncorrelated with previous methods, and are easily interpretable.

Abstract: A method of generating a model for performing account-level fraud detection is disclosed, and includes obtaining a plurality of candidate rules for referrals of financial transactions, calculating a savings density for each of the plurality of rules, selecting a rule from the plurality of rules having the maximum savings density, iteratively rebalancing savings densities for the remaining rules of the plurality of rules and selecting a next rule until a volume constraint is satisfied, and applying the selected rules to one or more financial transactions.

Abstract: Techniques for compressing data transaction history profiles are disclosed. Such profiles can include a plurality of profile variables with each profile variable comprising a real number that provides a factor for determining whether a proposed data transaction is indicative of fraud. A bit string is generated for each profile variable in the profiles that characterizes a first value plus a second value. The first value is equal to a mantissa for the real number corresponding to the profile variable. The second value is equal to a number of orders of magnitude above a minimum required expressed as multiples of the number of orders magnitude required to represent the plurality of real numbers in each the plurality of transaction history profiles divided by a range of bits. The generated bit string is stored as compressed profile variable within the data transaction history profiles. Related systems, apparatus, methods, and/or articles are also described.

Abstract: A method and apparatus comprising a fast and highly effective stochastic algorithm, referred to as Simmered Greedy Optimization (SG(N)), for solving combinatorial optimization problems, including the co-clustering problem comprising simultaneously clustering two finite sets by maximizing the mutual information between the clusterings and deriving maximally predictive feature sets. Co-clustering has found application in many areas, particularly statistical natural language processing and bio-informatics. Provided are results of tests on a suite of statistical natural language problems comparing SG(N) with simulated annealing and a publicly available implementation of co-clustering, wherein using SG(N) provided superior results with far less computation.