Abstract: Techniques for training a prediction model are disclosed. An example method includes processing historical event data comprising a plurality of computer-readable events for a resource to determine previous parameters for the resource. The method also includes generating training data for training the prediction model without using a forecast for future utilization of the resource. The training data comprises a set of proxies generated from previous parameters for the resource. Each proxy is associated with a remaining capacity of the resource and a remaining time to expiration of the resource. The method also includes training the prediction model to generate a mapping from the remaining capacity of the resource and the remaining time to expiration of the resource to the proxy. The method also includes receiving a request that describes a potential future event pertaining to the resource and generating a prediction for the potential future event using the prediction model.

Abstract: The technology evaluates the compliance of an AI application with predefined vector constraints. The technology employs multiple specialized models trained to identify specific types of non-compliance with the vector constraints within AI-generated responses. One or more models evaluate the existence of certain patterns within responses generated by an AI model by analyzing the representation of the attributes within the responses. Additionally, one or more models can identify vector representations of alphanumeric characters in the AI model's response by assessing the alphanumeric character's proximate locations, frequency, and/or associations with other alphanumeric characters. Moreover, one or more models can determine indicators of vector alignment between the vector representations of the AI model's response and the vector representations of the predetermined characters by measuring differences in the direction or magnitude of the vector representations.

Abstract: The technology evaluates the compliance of an AI application with predefined vector constraints. The technology employs multiple specialized models trained to identify specific types of non-compliance with the vector constraints within AI-generated responses. One or more models evaluate the existence of certain patterns within responses generated by an AI model by analyzing the representation of the attributes within the responses. Additionally, one or more models can identify vector representations of alphanumeric characters in the AI model's response by assessing the alphanumeric character's proximate locations, frequency, and/or associations with other alphanumeric characters. Moreover, one or more models can determine indicators of vector alignment between the vector representations of the AI model's response and the vector representations of the predetermined characters by measuring differences in the direction or magnitude of the vector representations.

Abstract: In systems for interpreting the predictions of a machine learning model with the help of a surrogate model, feature vectors of inputs to the machine learning model can be grouped based on locality sensitive hashes or other hashes that reflect similarity between the feature vectors in matching hash values. For a given prediction to be interpreted and the corresponding input feature vector, a suitable training dataset for the surrogate model can then be obtained at low computational cost by hashing the input feature vector and retrieving stored feature vectors with matching hash values, along with their respective predictions.

Abstract: Certain embodiments involve generating or optimizing a neural network for generating analytical or predictive outputs. The neural network can be generated using a relationship between various predictor variables and an outcome (e.g., a condition's presence or absence). The neural network can be used to determine a relationship between each of the predictor variables and a response variable. The neural network can be optimized by iteratively adjusting the neural network such that a monotonic relationship exists between each of the predictor variables and the response variable. The optimized neural network can be used both for accurately determining response variables using predictor variables and determining adverse action codes for the predictor variables, which indicate an effect or an amount of impact that a given predictor variable has on the response variable.

Abstract: Methods, computer readable media, and devices for machine learning for automated tagging of requests for e-commerce conversion funnel tracking. One method may include retrieving a plurality of requests previously grouped into a user session, generating preprocessed request information by determining a request type, a relative time, and a keyword count for each of the plurality of requests and determining a determining a relative time distribution for each of the plurality of keywords, and applying a machine learning model to the plurality of requests and the preprocessed request information to tag each of the plurality of requests as being associated with one of the plurality of conversion classifications.

Abstract: An apparatus for creating and training an artificial neural network (NN) performs iterations in which it implements and tests a version of a NN model, analyses the performance of the current NN model, generates a new NN module for improving the performance of the current NN model, adds the new NN module to the structure of the existing NN model to create a new NN model for executing, testing and analysing in the next iteration. In this way, the apparatus combines the creation and the testing of the NN model thereby improving the overall efficiency and the effectiveness of the development of NN based AI systems.

Abstract: Data-dependent node-to-node knowledge sharing to increase the interpretability of the activation pattern of one or more nodes in a neural network, is implemented by a set of knowledge sharing links. Each link may comprise a knowledge providing node or other source P and a knowledge receiving node R. A knowledge sharing link can impose a node-specific regularization on the knowledge receiving node R to help guide the knowledge receiving node R to have an activation pattern that is more easily interpreted. The specification and training of the knowledge sharing links may be controlled by a cooperative human-AI learning supervisor system in which a human and an artificial intelligence system work cooperatively to improve the interpretability and performance of the client system.

Abstract: A method including: determining a data type to be used for each of a plurality of operations, the plurality of operations being a series of operations to be used in machine learning processing; reducing a total processing time by modifying the determined data type to be used in each of the operations to a data type that is more accurate than the data type defined by the determining, the total processing time being a time period including an operation time for each of the plurality of operations and a conversion time taken to convert the data type; and executing each of the plurality of operations by using the modified data type of each of the plurality of operations.

Abstract: The systems and methods disclosed herein receive alphanumeric characters defining operative boundaries for expected model use cases, along with operational data. The expected model use cases share common attributes, which are used by a first AI model to construct observed model use cases from the operational data. Each observed model use case includes features such as a text-based description, expected input and output, AI model(s) generating the expected output from the input, and/or data supporting the AI models. For each observed model use case, a second AI model maps the alphanumeric characters and features to a risk category, selecting from multiple risk categories based on the level of risk associated with the features. The system identifies criteria for the observed model use case within the alphanumeric characters and generates gaps by comparing the criteria with the features of the observed model use case.

Abstract: A system and method for dynamic model training of a predictive machine learning model accesses data points of a training dataset including a plurality of model covariates. The predictive machine learning model is configured to generate an output including a risk rank representative of a mortality risk. The method selects one of the covariates and generates a historical data distribution for the selected covariate by applying the model to the training dataset including a plurality of historical application records. The method determines a current data distribution for the selected covariate. When comparison of the current data distribution with the historical data distribution indicates a data distribution shift exceeding a predetermined threshold, the method automatically updates parameters of the predictive machine learning model and retrains the predictive machine learning model using the updated parameters.

Abstract: The present disclosure relates to a method for performing inference on input data using a neural network and a processing device employing the aforementioned method. The method comprises the steps of obtaining and storing input data, obtaining parameter data indicating the parameters of the first layer and storing the parameter data in a parameter data storage location and processing the input data using the first layer parameter data, to form first layer output data. The method further comprises storing the first layer output data, obtaining parameter data of the second layer and storing the second layer parameter data by replacing the first layer parameter data with the second layer parameter data, processing the first layer output data using the stored second layer parameter data to form second layer output data; and storing the second layer output data.

Abstract: Various embodiments are generally directed to techniques for training deep neural networks, such as with an iterative approach, for instance. Some embodiments are particularly directed to a deep neural network (DNN) training system that generates a hardened DNN by iteratively training DNNs with images that were misclassified by previous iterations of the DNN. One or more embodiments, for example, may include logic to generate an adversarial image that is misclassified by a first DNN that was previously trained with a set of sample images. In some embodiments, the logic may determine a second training set that includes the adversarial image that was misclassified by the first DNN and the first training set of one or more sample images. The second training set may be used to train a second DNN. In various embodiments, the above process may be repeated for a predetermined number of iterations to produce a hardened DNN.

Abstract: A data storage device includes a memory array for storing data; a host interface for providing an interface with a host computer running an application; a central control unit configured to receive a command in a submission queue from the application and initiate a search process in response to a search query command; a preprocessor configured to reformat data contained in the search query command and generate a reformatted data; and one or more data processing units configured to extract one or more features from the reformatted data and perform a data operation on the data stored in the memory array in response to the search query command and return matching data from the data stored in the memory array to the application via the host interface.

Abstract: A data processing method used in neural network computing is provided. During a training phase of a neural network model, a feedforward procedure based on a calibration data is performed to obtain distribution information of a feedforward result for at least one layer of the neural network model. During the training phase of the neural network model, a bit upper bound of a partial sum is generated based on the distribution information of the feedforward result. During an inference phase of the neural network model, a bit-number reducing process is performed on an original operation result of an input data and a weight for the neural network model according to the bit upper bound of the partial sum to obtain an adjusted operation result.

Abstract: An exploratory recommender method and system generates affinity vectors from usage behaviors that are associated with stored content objects. A user's inferred content preferences are embodied within a multidimensional affinity space that is generated in accordance with the affinity vectors. An affinity space exploration algorithm explores areas within the multidimensional affinity space that are outside the user's inferred content preferences. The affinity space exploration algorithm is applied in accordance with the user's inferred content preferences to select a content object from the stored content objects. The content object is provided to the user.

Abstract: A quantum annealing schedule for a computational problem can be adjusted by methods and systems involving one or more processors. The one or more processors proceed by receiving a representation of the computation problem, the representation including a plurality of problem values. These problem values are transformed based on a plurality of trained parameters of a machine learning model to generate at least a portion of an annealing schedule including at least one annealing parameter. Instructions are transmitted to the quantum processor to cause the quantum processor to evolve from an initial state to a final state based on the computational problem and the at least a portion of an annealing schedule, the final state producing a result for the computational problem.

Abstract: A system and a method for increasing the classification confidence, with lesser dependence on large sets of training data, obtained by one or more machine learning based algorithms, by analyzing unstructured information using unstructured analysis pipeline comprising a probabilistic network such as a Bayesian network. The probabilistic network may comprise nodes associated with elements and cues defined by experts, and require fewer labelled data samples to train. The confidence level of the elements may be determined by machine learning and unstructured analysis methods and processed by the probabilistic network to estimate the confidence for a characterization quantity.

Abstract: A vector-based search method and system generates affinity vectors within a multidimensional affinity space by applying neural networks to stored content objects. A neural network is applied to content that is input by a user to generate an affinity vector, which is then compared to the affinity vectors corresponding to the stored content objects using a mathematical-based vector comparison algorithm. One or more of the stored content objects are selected and provided to the user based on the comparison. The selecting of the one or more stored content objects may be further performed based on an inference of a preference from user behaviors.

Abstract: A machine learning system, in particular a deep neural network. The machine learning system includes a plurality of layers that are connected to one another. The layers each ascertain an output variable as a function of an input variable and at least one parameter that is stored in a memory. The parameters of those layers that are connected to a further, in particular preceding, layer are each stored in the memory using a higher resolution than the parameters of those layers that are connected to a plurality of further, in particular preceding, layers. In addition, A method, a computer program, and a device for creating the machine learning system, are described.