Abstract: The described technology provides detection of a regime shift in streaming data by generating a curve fit of the streaming data representing an attribute over a duration to each of a plurality of probability density functions, scoring conformity of each curve fit to yield a plurality of fit scores, selecting a first duration probability density function among the plurality of probability density functions based on satisfaction of a probability density function fit condition by the fit score corresponding to the first duration probability density function, determining a probability density function change between the selected first duration probability density function and a second duration probability density function, wherein the second duration probability density function is selected for streaming data representing the attribute over a different duration, and indicating detection of the regime shift based on the determined probability density function change satisfying a shift condition.

Abstract: A sample of data, including a risk factor, is selected by a machine learning (ML) model of an extreme value theory (EVT) mechanism. A threshold is determined by the ML model based on the risk factor, an outlier score is generated for the sample, and the outlier score is compared to the threshold. The sample is identified as anomalous based on the generated outlier score being greater than the threshold. A schema comprising results of an investigation into the sample and the risk factor is updated based on the received schema.

Abstract: Techniques are described herein that are capable of providing privacy-preserving rules-based targeting using machine learning. Ranks are assigned to entities using a machine learning model. Values of each targetable feature associated with the respective entities are ordered. For each targetable feature, the entities are sorted among bins based on the values of the feature associated with the respective entities. For each targetable feature, a bin is selected from the bins that are associated with the feature based on the selected bin including more entities having respective ranks that are within a designated range than each of the other bins that are associated with the feature. A targeting rule is established, indicating a prerequisite for targeting an entity. The prerequisite indicating that the value of each targetable feature associated with the entity is included in a respective interval associated with the selected bin for the feature.

Abstract: Techniques are described herein that are capable of performing transfer-learning for structured data with regard to journeys defined by sets of actions. A first deep neural network (DNN) for a first journey is trained using structured data. Weights of nodes in the first DNN are transferred to nodes in a second DNN for a second journey using transfer-learning. An embedding layer replaces a final layer of the first DNN in the second DNN to provide an output with a same number of nodes as a pre-final layer of the first DNN. Weights of the nodes in the embedding layer are initialized based at least on a probability that a new feature of the second journey co-occurs with each feature in the structured data. A softmax function is applied on a final layer of the second DNN to indicate possible next actions of the second journey.

Abstract: A subset of data that includes a feature may be selected from a dataset. Parameters from the selected subset of data are determined and an extreme value theory (EVT) algorithm is implemented to determine a probability value for the feature based at least in part on the determined parameters. Based on the determined probability value for the feature, an outlier score is generated for the feature. Based on the outlier score being above a threshold, the subset is identified as anomalous.

Abstract: Techniques are described herein that are capable of generating a machine learning model based on constrained decision trees using a judgmental sample and feature ranking. A judgmental sample including observations, which include respective subsets of features, is generated. The observations are selected using multivariate stratified sampling. Important subsets of the features are determined based on each important subset being designated as more important than the other features by a respective individual. A score is determined for each feature, indicating a proportion of the important subsets that includes the respective feature. A highest scored feature is identified. Constrained decision trees having respective first splits are generated, based on respective subsets of the observations. A proportion of the first splits corresponding to the highest scored feature is based at least on the score of the highest scored feature.

Abstract: Methods, systems, apparatuses, and computer-readable storage mediums are descried for identifying a similarity between queries. An intermediate representation generator receives a set of queries from a repository, each query in the set of queries having generated a corresponding set of data stored in a data store. An intermediate representation is generated for each query, where the intermediate representation is characterized by a feature associated with text specified in the query. A similarity determiner determines similarity scores between pairs of intermediate representations. A pair of intermediate representations with a similarity score above a threshold is identified. An indication is generated that sets of data corresponding to queries corresponding to the intermediate representations are overlapping.

Abstract: Described are examples for generating a model for forecasting time series data. For a timeseries data set, one or more layers can be provided, where each layer in the one or more layers includes, for each timeseries data input in at least a portion of multiple timeseries data inputs, generating, for the timeseries data input, a short range output from a causal convolution process that is based on timeseries data inputs from the timeseries data set that are associated with timestamps within a threshold time before the timestamp of the timeseries data input, and generating, for the timeseries data input, a long range output from a transformer process based on the short range outputs from the causal convolution process for each timeseries data input from at least the portion of the multiple timeseries data inputs that are associated with timestamps before the timestamp of the timeseries data input.

Abstract: Described are examples for information discovery using domain-specific term matching, including obtaining, for a search session related to a domain, a set of search strings used for searching the domain for information discovery during the search session, identifying, for the search session, a string of the set of search strings as a search string hit and the remaining strings in the set of search strings as search string misses, and correlating, into a set of domain-specific search string misses, the remaining strings in the set of search strings with additional remaining search strings from one or more other search session related to the domain that have the same string identified as the search string hit.

Abstract: A system and method for recommending demand-supply agent pairs for transactions uses a deep neural network on data of demand agents to produce a demand agent vector, which is used to select supply agents based on their likelihood of future transaction and to find k nearest neighbor demand agents for each of the demand agents. The candidate supply agents and the k nearest neighbor demand agents are then combined to produce candidate demand-supply agent pairs, which are used to find recommended demand-supply agent pairs by applying modeling using machine learning.

Abstract: Example aspects include techniques for anomaly detection via sparse judgmental samples. These techniques may include determining, for an observation, anomalous autoencoder output (AAO) of a first layer of a first autoencoder trained used anomalous observations and determining, for the observation, conforming autoencoder output (CAO) of a second layer of a second autoencoder trained used conforming observations. In addition, the techniques may include calculating an anomaly value based on comparing the AAO to an anomalous signature output by the first layer of the first autoencoder, and calculating a conforming value based on comparing the CAO to a conforming signature output by the second layer of the second autoencoder. Further, the techniques may include determining whether the observation is anomalous based on comparing the anomaly value to the conforming value.

Abstract: Example aspects include techniques for anomaly detection via sparse judgmental samples. These techniques may include generating, via lexical analysis, a plurality of tokens from a textual representation of a machine learning (ML) model and generating, via a parser, based on the plurality of tokens, an abstract syntax tree (AST) corresponding to the ML model. In addition, the techniques may include identifying a data dependency of the ML model based on an AST node within the AST, the AST node corresponding to a data source and the data dependency indicating the ML model depends on the data source. Further, the techniques may include detecting a potential issue associated with the data source, and transmitting, based on the data dependency, an alert notification in response to the potential issue.

Abstract: A machine learning based monitoring focus engine is provided. Numeric and text features are collected from a computing system(s) and are utilized to determine if the system(s) will continue to run without issues or failures. That is, external characteristic information is received that corresponds to a predicted likelihood of a state that is associated with a processing system, and textual and numerical portions of the external characteristic information are mapped to neural network inputs. Word embedding is performed on the textual portion to generate embedded text features, and a plurality of inputs are provided to the neural network, where the plurality of inputs includes at least embedded text features, numerical features based on the numerical portion, and local features based on local characteristic information. Accordingly, the predicted likelihood of the state is determined based at least on an output of the neural network from the plurality of inputs.

Abstract: Systems and method for deep reinforcement learning are provided. The method includes generating, by a first neural network implemented on a processor, a synthetic data set based on an original data set, providing the original data set and the generated synthetic data set to a second neural network implemented on the processor, generating, by the second neural network, a prediction identifying the original data set and the generated synthetic data set, and based at least in part on the prediction incorrectly identifying the generated synthetic data set, exporting the generated synthetic data set.

Abstract: Systems and methods are described for identifying and resolving performance issues of automated components. The automated components are segmented into groups by applying a K-means clustering algorithm thereto based on segmentation feature values respectively associated therewith, wherein an initial set of centroids for the K-means clustering algorithm is selected by applying a set of context rules to the automated components. Then, for each group, a performance ranking is generated based at least on a set of performance feature values associated with each of the automated components in the group and a feature importance value for each of the performance features. The feature importance values are determined by training a machine learning based classification model to classify automated components into each of the groups, wherein the training is performed based on the respective performance feature values of the automated components and the respective groups to which they were assigned.

Abstract: Embodiments described herein are directed to ANN-based drift detection techniques for detecting data drift. For example, feature importance values of features provided to an ML model are determined. An input feature vector comprising a plurality of feature values are provided as an input to an autoencoder, which is configured to learn encodings representative of the features provided thereto and regenerate the features based on the encodings. The loss function (or re-construction loss) of the autoencoder is weighted by the feature importance values. A re-construction error based on the weighted loss is determined. The re-construction error is compared to a threshold condition. In response to determining that the re-construction error meets the threshold condition, a determination is made that the data has drifted. Responsive to determining that data has drifted, an action is taken with respect to the ML model to mitigate the data drift.

Abstract: A system and method for deduplicating target records using machine learning uses a deduplication machine learning model on the target records to classify the target records as duplicate target records and nonduplicate target records. The deduplication machine learning model leverages transfer learning, derived through first and second machine learning models for data matching, where the first machine learning model is trained using a generic dataset and the second machine learning model is trained using a target dataset and parameters transferred from the first machine learning model.

Abstract: A system and method for detecting non-compliances using machine learning uses anomaly detection on an input dataset of unlabeled observations to produce output observations with corresponding probability scores of the output observations being anomalous. A portion of the output observations are labeled as being compliant observations based on the corresponding probability scores, which are added to a training dataset of compliant and non-compliant observations to derive an augmented dataset of compliant and non-compliant observations. The augmented dataset of compliant and non-compliant observations is then used to train a machine learning model for non-compliance detection.

Abstract: Apparatus and method to detect anomalies in observations use a first plurality of observations regarding operation of a computing system, which are binned based on features values of the observations. Based on the binning, a weighting score is determined for the observations, which is applied to a loss function of an autoencoder. A second plurality of observations is then applied to the autoencoder as input to determine a reconstruction error value for each observation of the second plurality of observations. The reconstruction error values are used to detect anomalous observations of the second plurality of observations.

Abstract: A system and method for recommending demand-supply agent pairs for transactions uses a deep neural network on data of demand agents to produce a demand agent vector, which is used to select supply agents based on their likelihood of future transaction and to find k nearest neighbor demand agents for each of the demand agents. The candidate supply agents and the k nearest neighbor demand agents are then combined to produce candidate demand-supply agent pairs, which are used to find recommended demand-supply agent pairs by applying modeling using machine learning.