Embodiments of the present disclosure relate to feature selection via an ensemble of gating layers. According to embodiments of the present disclosure, a set of model parameter values for a machine learning model and a set of embedding vectors are determined for an input field of the machine learning model. The machine learning model is constructed to map an input sample in the input field to an embedding vector in the embedding vectors and process the embedding vector with the model parameter values to generate a model output. The machine learning model is trained by updating the model parameter values and the embedding vectors according to at least a first training objective function, the first training objective function being based on an orthogonality metric between embedding vectors in the embedding vectors and based on a difference between the model output and a ground-truth model output.

A device may receive data that is related to historical reports associated with an organization, historical audits of the historical reports, and individuals associated with the historical reports. The device may determine a multi-entity profile for the data. The multi-entity profile may include a set of groupings of the data by a set of attributes included in the data. The device may determine, using the multi-entity profile, a set of supervised model features for the historical reports. The device may determine, using the multi-entity profile, a set of unsupervised model features for the historical reports independent of the historical audits. The device may determine, utilizing a model, a score for a report. The device may perform one or more actions.

A data poisoning method and a data poisoning apparatus are provided. In the method, a training dataset and a validation dataset are retrieved. A perturbation is randomly initiated and added to data in the training dataset to generate poisoned training data. Values of multiple kernel functions of the poisoned training data and the validation dataset are computed by using kernel functions in a Gaussian process, and used to compute a mean of the Gaussian process on the validation dataset. A loss between the mean and the data in the validation dataset is computed by using a loss function of the Gaussian process, and used to generate an objective function that maximizes the loss. The objective function is solved to compute the perturbation that can maximize the loss.

The present disclosure relates to systems, methods, and computer readable media that evaluate performance of a machine learning system in connection with a test dataset. For example, systems disclosed herein may receive a test dataset and identify label information for the test dataset including feature information and ground truth data. The systems disclosed herein can compare the ground truth data and outputs generated by a machine learning system to evaluate performance of the machine learning system with respect to the test dataset. The systems disclosed herein may further generate feature clusters based on failed outputs and corresponding features and generate a number of performance views that illustrate performance of the machine learning system with respect to clustered groupings of the test dataset.

Various embodiments, methods and systems for implementing a distributed computing system machine-learning training service are provided. Initially a machine learning model is accessed. A plurality of synthetic data assets are accessed, where a synthetic data asset is associated with asset-variation parameters that are programmable for machine-learning. The machine learning model is retrained using the plurality of synthetic data assets. The machine-learning training service is further configured for executing real-time calls to generate an on-the-fly-generated synthetic data asset such that the on-the-fly-generated synthetic data asset is rendered in real-time to preclude pre-rendering and storing the on-the-fly-generated synthetic data asset. The machine-learning training service further supports hybrid-based machine learning training, where the machine learning model is trained based on a combination of the plurality of synthetic data assets, a plurality of non-synthetic data assets, and synthetic data asset metadata associated with the plurality of synthetic data assets.

A computer system is provided comprising a classification model management server computer configured, by instructions, to: receive a new image from a user device; apply a first digital model to first regions within the new image for classifying each of the first regions into a particular class; apply a second digital model to second regions within the new image for classifying each of the second regions into a particular class; and transmit classification data related to the class of the first regions and the class of the second regions to the user device. In connection therewith, the second regions each generally correspond to a combination of multiple first regions.

Methods and systems for training a machine learning model include training a machine learning model using training data. A status of the machine learning model's training is determined based on an accuracy curve of the machine learning model over the course of the training. Parameters of the training are adjusted based on the status. Training of the machine learning model is completed using the adjusted parameters.

An example operation may include one or more of generating, by a training participant client comprising a training dataset, a plurality of transaction proposals that each correspond to a training iteration for machine learning model training related to stochastic gradient descent, the machine learning model training comprising a plurality of training iterations, the transaction proposals comprising a gradient calculation performed by the training participant client, a batch from the private dataset, a loss function, and an original model parameter, receiving, by one or more endorser nodes of peers of a blockchain network, the plurality of transaction proposals, and evaluating each transaction proposal.

Virtual game worlds for computer games can be provided using machine learning. The use of machine learning enables the virtual game worlds to be generated at run time by standard consumer hardware devices. Machine learning agents are trained in advance to the characteristics of the particular game world. Then, these suitably trained machine learning agents can be used to generate a relevant portion of a virtual game world, such as a portion of the virtual game world that is proximate to a play's position. Advantageously, the virtual game world can be provided in high resolution and is able to cover a substantially larger region than conventional practical.

Interfaces and systems are provided for harvesting ground truth from forms to be used in training models based on key-value pairings in the forms and to later use the trained models to identify related key-value pairings in new forms. Initially, forms are identified and clustered to identify a subset of forms to label with the key-value pairings. Users provide input to identify keys to use in labeling and then select/highlight text from forms that are presented concurrently with the keys in order to associate the highlighted text with the key(s) as the corresponding key-value pairing(s). After labeling the forms with the key-value pairings, the key-value pairing data is used as ground truth for training a model to independently identify the key-value pairing(s) in new forms. Once trained, the model is used to identify the key-value pairing(s) in new forms.