Intuit, Inc.

Workday, Inc.

Efraim David Feinstein

Lucian Vlad Lita

Jonathan Goldman

Vladimir Giverts

Adeyemi Ajao

Kristy Gateley

Jan Prach

Tarek Rabbani

Anuranjita Tewary

Barzel Segal

Phil Monroe

Sricharan Kallur Palli Kumar

Conrad De Peuter

Preeti Duraipandian

Sumanth Venkatasubbaiah

Manas Kumar Mukherjee

Yi Xu Ng

Nagaraj Janardhana

Riley F Edmunds

Fadoua Khmaissia

Ian Andrew Sebanja

The present disclosure provides techniques for schema-based machine learning model monitoring. One example method includes receiving input data to and output data related to a machine learning model, wherein the input data and the output data conform to a data schema, retrieving, based on the data schema, a set of fields associated with the input data and the output data, performing statistical analysis for the machine learning model based on the set of fields retrieved, and predicting one or more attributes of the machine learning model based on the statistical analysis, wherein the one or more attributes of the machine learning model indicate a result of monitoring of the machine learning model, explainability information related to the machine learning model, or health of the machine learning model.

Schema-based machine learning model monitoring

A method that includes extracting image features of a document image, executing an optical character recognition (OCR) engine on the document image to obtain OCR output, and extracting OCR features from the OCR output. The method further includes executing an anomaly detection model using features including the OCR features and the image features to generate anomaly score, and presenting anomaly score.

Multiple input machine learning framework for anomaly detection

A method and system distinguish between anomalous members of a majority group and members of a target group. The system and method utilize a neural network architecture that attends to each level of a classification hierarchy. The system and method chain a semi-supervised autoencoder with a supervised classifier neural network. The autoencoder is trained in a semi-supervised manner with a machine learning process to identify user profile data that are typical of a majority class. The classifier neural network is trained in a supervised manner with a machine learning process to distinguish between user profile data that are anomalous members of the majority class and user profile data that are members of the target class.

System and method for hierarchical deep semi-supervised embeddings for dynamic targeted anomaly detection

Aspects of the present disclosure provide techniques for machine learning and rules integration. Embodiments include receiving input values corresponding to a subset of a set of input variables associated with an automated determination. Embodiments include generating a directed acyclic graph (DAG) representing a set of constraints corresponding to the set of input variables. The set of constraints relate to one or more machine learning models and one or more rules. Embodiments include receiving one or more outputs from the one or more machine learning models based on one or more of the input values. Embodiments include determining outcomes for the one or more rules based on at least one of the input values. Embodiments include populating the DAG based on the input values, the one or more outputs, and the outcomes. Embodiments include making the automated determination based on logic represented by the DAG.

Integrated machine learning and rules platform for improved accuracy and root cause analysis

A method includes establishing a network connection with a source computing device and an application services computing device, receiving, via the network connection, a source event stream at the application services computing device, and extracting a sample of the source event stream. The method further includes partitioning the sample of the source event stream into fields, identifying a field data type of a field of the multiple fields in the sample, identifying a distribution of values of the field in the sample, and extrapolating, from the sample of the source event stream, extrapolated functions for the fields. Extrapolating an extrapolated function is dependent on the field data type and the distribution of the field. The method further includes transforming, based on the plurality of extrapolated functions in the configuration file, the source event stream to obtain a transformed event stream, and analyzing, by a target machine learning model, the transformed event stream.

Generic event stream processing for machine learning

A system for determining a canonical title includes an interface and a processor. The interface is to receive a new title. The processor is to generate a set of n-grams based on the new title; look up n-grams in canonical database to assign one or more labels; and map the new title to a canonical title based on the one or more labels.

Method for systematic mass normalization of titles

A method for normalizing raw titles to canonical titles is described. The method includes designating a set of canonical titles, generating a set of n-grams for each canonical title, assigning a set of attributes to each n-gram, assigning a set of labels to each of the attributes, and storing the labeled canonical title and labeled n-grams in a database. In some examples, a new title may be mapped to an existing canonical title in the database by generating a set of n-grams for the new title, looking up the n-grams in the database of canonical titles, retrieving the set of labels assigned to n-grams in the database that match n-grams from the new title, and assigning those labels to the corresponding attributes of the new title. The new title may then be mapped to a canonical title on the basis of similarly labeled attributes.

Growing community of inventors

San Jose, CA, United States of America

Efraim David Feinstein