Abstract: Systems and methods for training machine learning models are disclosed. An example method includes receiving a semi-labeled set of training samples including a first set of training samples, where each training sample in the first set is assigned a known label, and a second set of training samples, where each training sample in the second set has an unknown label, determining a first loss component, the first loss component providing a loss associated with the first set, determining a second loss component, the second loss component having a value which increases based on a difference between a distribution of individually predicted values of at least the second set and an expected overall distribution of at least the second set, and training the machine learning model, based on the first loss component and the second loss component, to predict labels for unlabeled input data.

Abstract: A processor may identify a plurality of data sets subject to upcoming update processing in a next update cycle. For each of the plurality of data sets, the processor may determine a probability that data included in the data set has changed since a most recent update processing. The processor may exclude a first subset of the plurality of data sets having respective probabilities below a threshold value from the upcoming update processing until the respective probabilities are determined again in a subsequent update cycle. The processor may perform the upcoming update processing on the plurality of the data sets not included in the first subset, where the upcoming update processing may include obtaining updated data from at least one external data source.

Abstract: A method and system that proactively generate alerts for updating a scraping script to avoid scraping script errors. A predetermined number of webpages targeted by the scraping script are randomly sampled. The scraping script is appended to each webpage in the sample. A structured list of text fragments across the webpages with the appended script is generated. At predetermined time intervals, a fresh set of webpages is sampled, the scraping script is appended to the webpages, and a new structured list is generated. If the new structured list and the previous structured list do not match, the webpages may have been changed and the scraping script may have to be updated. An alert is generated indicating that such update is required and may include a location of the mismatch. Therefore, scraping script errors are proactively detected and can be rectified before an actual error occurs and propagates.

Abstract: Certain aspects of the present disclosure provide techniques for medoid-based data compression. One example method generally includes receiving item data indicative of one or more items, determining one or more medoids based on the item data, determining, for each item of the one or more items, a corresponding medoid based on the one or more medoids, identifying, for each item of the one or more items, a difference between the item and the corresponding medoid for the item, storing the one or more medoids, and storing, for each item of the one or more items, the identified difference between the item and the corresponding medoid.

Abstract: Methods and systems are presented for imputing missing data items within a first dataset based on data associated with a second dataset that is the nearest neighbor of the first dataset. A first mapping model is configured to map data subsets corresponding to a first data source to first positions in a multi-dimensional space. A second mapping model is configured to map data subsets corresponding to a second data source to second positions in the multi-dimensional space. The first and second mapping models are trained together to reduce a distance between positions mapped by the first and second mapping models based on corresponding data subsets that belong to the same entity. A nearest neighbor dataset to the first dataset is identified based on the first and second mapping models. Data associated with the nearest neighbor dataset is used to impute the missing data items of the first dataset.

Abstract: Certain aspects of the present disclosure provide techniques for medoid-based data compression. One example method generally includes receiving item data indicative of one or more items, determining one or more medoids based on the item data, determining, for each item of the one or more items, a corresponding medoid based on the one or more medoids, identifying, for each item of the one or more items, a difference between the item and the corresponding medoid for the item, storing the one or more medoids, and storing, for each item of the one or more items, the identified difference between the item and the corresponding medoid.

Abstract: Computer system security and efficiency of processing operations may be improved using techniques that are described relating to analyzing user actions based on time of day of occurrence, and using time of a day as a factor in determining whether a particular action should be allowed or disallowed. Past action times can be transformed to a two-dimensional representation using a radial time schema that avoids discontinuity. A probability distribution can indicate a likelihood of whether a new action fits a previous pattern. If a new user action is relatively unlikely due to time of day, the new user action might be denied/prevented from completing, thus enhancing computer system security and avoiding unnecessary computational processing costs.

Abstract: Systems and methods for generating a contextually adaptable classifier model are disclosed. An example method is performed by one or more processors of a system and includes obtaining a dataset, feature values, and labels, transforming each datapoint into a natural language statement (NLS) associating the datapoint's feature values and label with feature identifiers and a label identifier, generating a feature matrix for each NLS, transforming the feature matrix into a global feature vector, generating a target vector for each NLS, transforming the target vector into a global target vector having a same shape, and generating, using the vectors, a similarity measurement operation, and a loss function, a classifier model trained to generate a compatibility score predictive of an accuracy at which the classifier model can classify given data based on at least one of a different feature characterizing the given data or a different label for classifying the given data.

Abstract: Techniques are disclosed relating to weakly supervised machine learning, which may be employed when there is a limited amount of labeled data available. A computer system may generate respective sets of synthetic labels for unlabeled data for a classification problem, where a given set of synthetic labels is produced by a corresponding one of a plurality of different label models. The computer system may then fit a set of supervised models, where each supervised model is fitted with one of the respective sets of synthetic labels to produce a respective set of predictions. The computer system may then evaluate the set of supervised models based on their respective set of predictions and using a set of labeled data for the classification problem. The evaluation may be used to select a particular supervised model and its corresponding label model.

Abstract: A system deploying a machine learning technique that utilizes known code graph and abstract syntax tree pairs for known JSON objects to learn a function for predicting a corresponding abstract syntax tree from a new JSON object. The predicted abstract syntax tree is used to generate code for formatting the new JSON object into a standardized data structure.

Abstract: Systems and methods for query-based recommendation systems using machine learning-trained classifiers are provided. A service provider server receives, from a communication device through an application programming interface, a query in an interaction between the server provider server and the communication device. The service provider server generates a vector of first latent features from a set of first visible features associated with the query using a machine learning-trained classifier. The service provider server generates a likelihood scalar value indicating a likelihood of the query is answered by a candidate user in a set of users using a combination of the vector of first latent features and a vector of second latent features. The service provider server provides, to the communication device through the application programming interface, a recommendation message as a response to the query, where the recommendation message includes the likelihood scalar value and an indication of the candidate user.

Abstract: The present disclosure provides techniques for extracting structural information using machine learning. One example method includes receiving electronic data indicating one or more pages, constructing, for each page of the one or more pages, a tree based on the page, wherein each level of the tree includes one or more nodes corresponding to elements in a level of elements in the page, encoding, for each page of the one or more pages, a value of each node of the tree for the page into a vector using a first machine learning model, sampling a plurality of pairs of vectors from the one or more trees for the one or more pages, wherein a given pair of vectors corresponds to values of nodes in a same tree, training a second machine learning model using the plurality of pairs, and combining each vector with weights of the second machine learning model.

Abstract: Methods and systems are presented for imputing missing data items within a first dataset based on data associated with a second dataset that is the nearest neighbor of the first dataset. A first mapping model is configured to map data subsets corresponding to a first data source to first positions in a multi-dimensional space. A second mapping model is configured to map data subsets corresponding to a second data source to second positions in the multi-dimensional space. The first and second mapping models are trained together to reduce a distance between positions mapped by the first and second mapping models based on corresponding data subsets that belong to the same entity. A nearest neighbor dataset to the first dataset is identified based on the first and second mapping models. Data associated with the nearest neighbor dataset is used to impute the missing data items of the first dataset.

Abstract: Methods and systems are presented for providing a framework that configures a machine learning model to be insensitive to changes in input features. A computer modeling system determines data sources from which attribute values associated with transactions can be obtained. Instead of configuring the machine learning model to accept the attribute values as inputs, the computer modeling system may configure the machine learning model to accept a vector representation in a multi-dimensional space as input values. The computer modeling system then generates an encoder for each data source. Each encoder is configured to encode attribute values from a corresponding data source to a representation representing the attribute values. Further, each encoder is trained to minimize a variance between outputs of the different encoders. The computer modeling system determines a vector representation based on the representations generated by the encoders and provide the vector representation to the machine learning model.

Abstract: Techniques for detecting fraud may include mapping routing numbers of one or more financial institutions with geolocation data of the financial institutions; obtaining a geolocation of a user based on the user's internet protocol (IP) address; obtaining a first user input from the user indicating a first financial institution; generating a match score for each of the one or more financial institutions that indicates a level of match between the first user input and the respective financial institution; boosting the match score for each financial institution based on its location with respect to the geolocation of the user; generating a list of financial institutions having the boosted match score above a threshold; obtaining a second user input from the user indicating at least one second financial institution; and presenting search results to the user based on the second user input, wherein the search results are boosted.

Abstract: Methods and systems for efficiently generating tagged training data for machine learning models. In conventional systems, all of the raw data (e.g., each sentence) has to be manually tagged. Instead, the methods and systems generate a representative sample for multiple portions of raw data, e.g., a representative sentence for multiple, similar sentences. Only the representative sample is tagged and used for training, thereby realizing a significant efficiency in both tagging the data and training the machine learning models.

Abstract: A computing device generates a first token for first data content that is associated with a first relationship and a second relationship, and a second token for second data content that is associated with the first relationship and a third relationship, such that the first token and second token are generated based on a frequency of use of data values included in the first and the second data content. The computing device calculates a first similarity score of data values from third data content that is associated with the second relationship and a fourth relationship with data values from fourth data content that is associated with the third relationship and the fourth relationship in response to the first and second token matching. The computing device then performs, in response to the first similarity score satisfying a similarity threshold, a first modification to any of the data content.

Abstract: Methods and systems for providing a machine learning model that can perform predictions based on incomplete input values are presented. The machine learning model includes multiple input layers of input nodes, where input nodes from different input layers can be connected with each other. Based on the connections among the input nodes, certain input values can be inferred from other input values. When a request is received, it is determined which input values are available and which input values are missing. Based on which input values are available, the machine learning model is modified by masking a subset of connections among nodes in the input layers. The modified machine learning model is then configured to infer the missing input values from the available input values, and to provide an output based on the available input values and the inferred input values. The request is processed based on the output.

Abstract: Techniques are disclosed relating to assessing technology activity using image-based machine learning algorithms. A computer system may access a data set that includes a plurality of parameters (e.g., technologies) for an item (e.g., a web-based interface). The plurality of parameters may correspond to a plurality of time intervals. The computer system may generate a two-dimensional graphical representation of the data set. A first dimension of the graphical representation may be indicative of values of the plurality of parameters at different time intervals and a second dimension of the graphical representation may be indicative of a time period that includes the plurality of time intervals. At least one characteristic of the data set may be determined by inputting the graphical representation of the data set to a trained machine learning module. The trained machine learning module may implement an image-based learning algorithm.

Abstract: Various techniques for determining risk assessment predictions and decisions are disclosed. Certain disclosed techniques include the implementation of decision-tree based models in determining predictions of risk for an operation based on an input dataset. The disclosed techniques include pruning decision trees to compensate for deprecation of variables from the input dataset. Decision trees may be pruned at nodes associated with the deprecated variables to inhibit the decision trees from breaking down during operation on an input dataset having deprecated variables.