Abstract: An object-extraction method includes generating multiple partition objects based on an electronic document, and receiving a first user selection of a data element via a user interface of a compute device. In response to the first user selection, and using a machine learning model, a first subset of partition objects from the multiple partition objects is detected and displayed via the user interface. A user interaction, via the user interface, with one of the partition objects is detected, and in response, a weight of the machine learning model is modified, to produce a modified machine learning model. A second user selection of the data element is received via the user interface, and in response and using the modified machine learning model, a second subset of partition objects from the multiple partition objects is detected and displayed via the user interface, the second subset different from the first subset.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for generating an output sequence from an input sequence. In one aspect, one of the systems includes an encoder neural network configured to receive the input sequence and generate encoded representations of the network inputs, the encoder neural network comprising a sequence of one or more encoder subnetworks, each encoder subnetwork configured to receive a respective encoder subnetwork input for each of the input positions and to generate a respective subnetwork output for each of the input positions, and each encoder subnetwork comprising: an encoder self-attention sub-layer that is configured to receive the subnetwork input for each of the input positions and, for each particular input position in the input order: apply an attention mechanism over the encoder subnetwork inputs using one or more queries derived from the encoder subnetwork input at the particular input position.

Abstract: A system, apparatus or process that applies artificial intelligence associated with improved modeling and validation techniques to identify a series of compatible components, e.g., to accomplish an analytical task. In connection with embodiments of the invention, an input module receives input data comprising an inquiry associated with accomplishing a task. A model module receives the input data and designs at least one pipeline comprising a plurality of components designed to accomplish the task. A compatibility module determines at least one valid pipeline by analyzing the at least one pipeline and determining whether each one of the plurality of components are compatible with a component immediately before and a component immediately after the one of the plurality of components. A display module displays the at least one valid pipeline.

Abstract: The embodiments herein disclose a system and method for implementing reinforcement learning agents using a reinforcement learning processor. An application-domain specific instruction set (ASI) for implementing reinforcement learning agents and reward functions is created. Further, instructions are created by including at least one of the reinforcement learning agent ID vectors, the reinforcement learning environment ID vectors, and length of vector as an operand. The reinforcement learning agent ID vectors and the reinforcement learning environment ID vectors are pointers to a base address of an operations memory. Further, at least one of said reinforcement learning agent ID vector and reinforcement learning environment ID vector is embedded into operations associated with the decoded instruction. The instructions retrieved by agent ID vector indexed operation are executed using a second processor, and applied onto a group of reinforcement learning agents.

Abstract: A computer trains a neural network model. (A) Observation vectors are randomly selected from a plurality of observation vectors. (B) A forward and backward propagation of a neural network is executed to compute a gradient vector and a weight vector. (C) A search direction vector is computed. (D) A step size value is computed. (E) An updated weight vector is computed. (F) Based on a predefined progress check frequency value, second observation vectors are randomly selected, a progress check objective function value is computed given the weight vector, the step size value, the search direction vector, and the second observation vectors, and based on an accuracy test, the mini-batch size value is updated. (G) (A) to (F) are repeated until a convergence parameter value indicates training of the neural network is complete. The weight vector for a next iteration is the computed updated weight vector.

Abstract: In general, certain embodiments of the present disclosure provide methods and systems for enabling a reproducible processing of machine learning models and scalable deployment on a distributed network. The method comprises building a machine learning model; training the machine learning model to produce a plurality of versions of the machine learning model; tracking the plurality of versions of the machine learning model to produce a change facilitator tool; sharing the change facilitator tool to one or more devices such that each device can reproduce the plurality of versions of the machine learning model; and generating a deployable version of the machine learning model through repeated training.

Abstract: A system for self-learning archival of electronic data may be provided. A binary classifier may identify a text segment of an electronic dataset in response to text of the electronic dataset being associated with indicators of a word model. A first multiclass classifier may generate a first classification set comprising respective statistical metrics for the datafield that each predefined identifier in a group of predefined identifiers is representative of the datafield. A second multiclass classifier may receive a context of the electronic dataset and generate a second classification set. A combination classifier may apply weight values to the first classification set and the second classification set and form a weighted classification set and select a predefined identifier as being representative of the datafield based on the weighted classification set. The processor may store, in a memory, a data record comprising an association between the predefined identifier and the datafield.

Abstract: The present disclosure provides a method and a device for optimization processing of neural network models. The method includes the following: determining one or more target layers of the neural network model based on the number of neurons at each layer of the neural network model; for each of the one or more target layers, adding a virtual layer between the target layer and a preceding layer of the target layer, where neurons at the virtual layer are separately connected to neurons at the target layer and neurons at the preceding layer of the target layer, and addition of the virtual layer reduces the number of connections between the target layer and the preceding layer of the target layer; and training the neural network model after having added the virtual layers, to obtain an optimized neural network model.

Abstract: Generating a computing specification to be executed by a quantum processor includes: accepting a problem specification that corresponds to a second-quantized representation of a fermionic Hamiltonian, and transforming the fermionic Hamiltonian into a first qubit Hamiltonian including a first set of qubits that encode a fermionic state specified by occupancy of spin orbitals. An occupancy of any spin orbital is encoded in a number of qubits that is logarithmic in the number of spin orbitals, and a parity for a transition between any two spin orbitals is encoded in a number of qubits that is logarithmic in the number of spin orbitals. An eigenspectrum of a second qubit Hamiltonian, including the first set of qubits and a second set of qubit, includes a low-energy subspace and a high-energy subspace, and an eigenspectrum of the first qubit Hamiltonian is approximated by a set of low-energy eigenvalues of the low-energy subspace.

Abstract: An apparatus for artificial intelligence acceleration is provided. The apparatus includes a storage and compute system having a distributed, redundant key value store for metadata. The storage and compute system having distributed compute resources configurable to access, through a plurality of authorities, data in the solid-state memory, run inference with a deep learning model, generate vectors for the data and store the vectors in the key value store.

Abstract: Techniques are described for generating predictive alerts. In one or more embodiments, a seasonal model is generated, the seasonal model representing one or more seasonal patterns within a first set of time-series data, the first set of time-series data comprising data points from a first range of time. A trend-based model is also generated to represent trending patterns within a second set of time-series data comprising data points from a second range of time that is different than the first range of time. A set of forecasted values is generated based on the seasonal model and the trend-based model. Responsive to determining that a set of alerting thresholds has been satisfied based on the set of forecasted values, an alert is generated.

Abstract: A method of mapping a first operation of a source framework to a second operation of a target framework for an artificial neural network includes determining an alignment between a current source axis order and a current target axis order. The method also includes setting the current target axis order based on the alignment, and an expected source axis order of the first operation and/or an expected target axis order of the second operation.

Abstract: Embodiments of the present invention disclose a method, computer program product, and system for an automated chat bot conversation session and an agent transfer system for the conversation session. The computer receives a user input from a user in an automated chat bot conversation session. The computer analyzes the user input for at least one sentiment, wherein an at least one analysis result is a value assigned to the at least one sentiment contained within the user input. The computer compares the at least one analysis result to a threshold value to determine if the user should be transferred from the automated chat bot conversation session to a conversation session with a suitable agent. The computer then transfers the user to the conversation session with the suitable agent.

Abstract: A system predicts user intent to take an action and delivers content items to the user that match that intent. A plurality of features or attributes for each tracking pixel in a set of tracking pixels can be acquired based on content items and landing pages associated with each tracking pixel. For example, features for a tracking pixel can be determined based on information associated with a content item that enabled a user to access a landing page from which the tracking pixel was fired or triggered. In this example, features for the tracking pixel can also be determined based on information associated with the landing page. The features for the tracking pixels can be utilized to train a machine learning model. The machine learning model can be trained to predict whether or not a particular user intends to produce a conversion (e.g., make a purchase).

Abstract: A first cognitive instance receives information about other cognitive instances and from this compiles a cognitive community map that associates individual ones of the other cognitive instances with specific capabilities of said respective other cognitive instances. The first cognitive instance stores that map in a local memory of the first cognitive instance; and when the first cognitive instance executes a cognitive computing program it checks the cognitive community map for at least one of the specific capabilities relevant for executing that program to address/satisfy a user request that caused the program to execute. In various embodiment these cognitive instances share their respective cognitive capabilities via cognitive capability maps, which may be refreshed in their local memories for example by sending a broadcast message. Thus any given cognitive instance can identify cognitive peers with the capabilities most relevant to assist itself in solving a given problem/request.

Abstract: A reinforcement learning processor specifically configured to train reinforcement learning agents in the AI systems by the way of implementing an application-specific instruction set is disclosed. The application-specific instruction set incorporates ‘Single Instruction Multiple Agents (SIMA)’ instructions. SIMA type instructions are specifically designed to be implemented simultaneously on a plurality of reinforcement learning agents which interact with corresponding reinforcement learning environments. The SIMA type instructions are specifically configured to receive either a reinforcement learning agent ID or a reinforcement learning environment ID as the operand. The reinforcement learning processor is designed for parallelism in reinforcement learning operations. The reinforcement learning processor executing of a plurality of threads associated with an operation or task in parallel.

Abstract: Systems and methods of evaluating rules. Other embodiments are also described.

Abstract: A machine learning device included in a cutting fluid supply timing control device observes operating state data regarding an operating state of a cutting fluid supply device as a state variable representing a current environment state, acquires supply timing data indicating a timing of supplying a cutting fluid as label data, and then learns the operating state data and the supply timing data in association with each other by using these state variable and label data.

Abstract: A novel data-parallel algorithm is presented for topic modeling on a highly-parallel hardware architectures. The algorithm is a Markov-Chain Monte Carlo algorithm used to estimate the parameters of the LDA topic model. This algorithm is based on a highly parallel partially-collapsed Gibbs sampler, but replaces a stochastic step that draws from a distribution with an optimization step that computes the mean of the distribution directly and deterministically. This algorithm is correct, it is statistically performant, and it is faster than state-of-the art algorithms because it can exploit the massive amounts of parallelism by processing the algorithm on a highly-parallel architecture, such as a GPU. Furthermore, the partially-collapsed Gibbs sampler converges about as fast as the collapsed Gibbs sampler and identifies solutions that are as good, or even better, as the collapsed Gibbs sampler.

Abstract: Systems and methods for automated sequencing database generation are disclosed herein. The system can include memory that can include a content library database; a graph database; and a model database. The system can include a user device and at least one server. The at least one server can: receive a content aggregation from the content library database; identify content components of the content aggregation based on a natural language processing analysis of at least a portion of the content aggregation; identify explicit sequencing of the content components; generate an intermediate content graph based on the explicit sequencing of the content components; generate a final content graph from the intermediate content graph based on implicit sequencing of the content components; and store the final content graph within the graph database.