Abstract: An application performance analyzer adapted to analyze the performance of one or more applications running on IT infrastructure, comprises: a data collection engine collecting performance metrics for one or more applications running on the IT infrastructure; an anomaly detection engine analyzing the performance metrics and detecting anomalies, i.e. performance metrics whose values deviate from historic values with a deviation that exceeds a predefined threshold; a correlation engine detecting dependencies between plural anomalies, and generating anomaly clusters, each anomaly cluster consisting of anomalies that are correlated through one or more of the dependencies; a ranking engine ranking anomalies within an anomaly cluster; and a source detection engine pinpointing a problem source from the lowest ranked anomaly in an anomaly cluster.

Abstract: Aspects of processing data for Long Short-Term Memory (LSTM) neural networks are described herein. The aspects may include one or more data buffer units configured to store previous output data at a previous timepoint, input data at a current timepoint, one or more weight values, and one more bias values. The aspects may further include multiple data processing units configured to parallelly calculate a portion of an output value at the current timepoint based on the previous output data at the previous timepoint, the input data at the current timepoint, the one or more weight values, and the one or more bias values.

Abstract: A method for assessing movement of a body portion includes, via one or more machine learning models, analyzing a sensor signal indicative of movement of the body portion to determine a movement of the body portion; determining a sensor confidence level based, at least in part, on a characteristic of the sensor signal; receiving a series of images indicative of movement of the body portion; measuring an angle of movement of the body portion; determining a vision confidence level based, at least in part, on a quality of an identification the body portion; selecting the sensor signal, the measured angle of movement, or a combination thereof as an input into a machine learning model based on the sensor confidence level and the vision confidence level, respectively; analyzing the input to determine a movement pattern of the body portion; and outputting the movement pattern to a user.

Abstract: A system and method for constructing training dictionaries with multichannel information. An exemplary method takes into account the effect of the acoustic path while training multichannel acoustic data. A method that uses different time-frequency resolutions in machine learning training is also presented.

Abstract: For a visit of a user to a web page where the user's identity on an online system is not presently known to the online system, the online system uses a machine learning model to make a prediction of the user's identity. The online system obtains visit data about the visit of the user to the web page. The online system identifies candidate user IDs that may represent the user, based on the visit data and data known about previous visits of the candidate user IDs. The online system derives visit features for each candidate user ID based on a relationship between the current visit data and previous visit data for the candidate user ID. The online system provides the visit features for each candidate user ID to a prediction model that determines whether, or how likely, the candidate user ID accurately identifies the visiting user, and based on the determinations selects one of the candidate user IDs as the most likely user ID for the visiting user.

Abstract: A neural network system comprises a plurality of neurons, comprising a layer of input neurons, one or more layers of hidden neurons, and a layer of output neurons. The system further comprises a plurality of arrays of weights, each array of weights being configured to receive a plurality of discrete data points from a first layer of neurons and to produce a corresponding discrete data point to a second layer of neurons during a feed forward operation, each array of weights comprising a plurality of resistive processing units (RPU) having respective settable resistances. The system includes a neuron control system configured to control an operation mode of each of the plurality of neurons, wherein the operation mode comprises: a feed forward mode, a back propagation mode, and a weight update mode.

Abstract: Methods and systems for selecting a selected-sub-set of features from a plurality of features for training a machine learning module, the training of the machine learning module to enable classification of an electronic document to a target label, the plurality of features associated with the electronic document. In one embodiment, the method comprises analyzing a given training document to extract the plurality of features, and for a given not-yet-selected feature of the plurality of features: generating a set of relevance parameters iteratively, generating a set of redundancy parameters iteratively and determining a feature significance score based on the set of relevance parameters and the set of redundancy parameters. The method further comprises selecting a feature associated with a highest value of the feature significance score and adding the selected feature to the selected-sub-set of features.

Abstract: An expert recommendation method, system, and non-transitory computer readable medium, include a topic extraction circuit configured to extract a topic of a user input message in real-time, an expert recommending circuit configured to recommend a list including a plurality of experts based on the extracted topic, and an expert ranking circuit configured to order the experts on the list of experts based on an expert rank factor.

Abstract: A method for transferring acoustic knowledge of a trained acoustic model (AM) to a neural network (NN) includes reading, into memory, the NN and the AM, the AM being trained with target domain data, and a set of training data including a set of phoneme data, the set of training data being data obtained from a domain different from a target domain for the target domain data, inputting training data from the set of training data into the AM, calculating one or more posterior probabilities of context-dependent states corresponding to phonemes in a phoneme class of a phoneme to which each frame in the training data belongs, and generating a posterior probability vector from the one or more posterior probabilities, as a soft label for the NN, and inputting the training data into the NN and updating the NN, using the soft label.

Abstract: In response to receiving a user inquiry describing one more symptoms of a medical condition, an online hospital visit prediction system can determine whether the user will visit the hospital for treatment related to the medical condition the day following the user inquiry, termed “hospital day.” Query logs are stored for a large plurality of users. Using the query logs, a prediction model can be trained using predetermined features extracted from the query logs of each of a plurality of users for a predetermined period of time prior to the day the user visited the hospital. The predetermined features can be used to train the prediction model with a high accuracy to determine whether a future user will visit the hospital the day following a query about a medical condition.

Abstract: A system and method for automating proactive communication. The information for the desired contacts may be accepted from a user. A selection of contact communication frequency preferences may be received from a user. An automatic communication to one of the desired contacts may be initiated. The user may be allowed to cancel the automatic communication, in response to receiving a notification that the communication is about to begin. A response indicative of the communication status may be received. Rules and preferences may be optimized based upon the received response.

Abstract: Systems, methods, and computer program products to provide neural embeddings of transaction data. A network graph of transaction data based on a plurality of transactions may be received. The network graph of transaction data may define relationships between the transactions, each transaction associated with at least a merchant and an account. A neural network may be trained based on training data comprising a plurality of positive entity pairs and a plurality of negative entity pairs. An embedding function may then encode transaction data for a first new transaction. An embeddings layer of the neural network may determine a vector for the first new transaction based on the encoded transaction data for the first new transaction. A similarity between the vectors for the transactions may be determined. The first new transaction may be determined to be related to the second transaction based on the similarity.

Abstract: A data entry system is described which has a user interface which receives a sequence of one or more context text items input by a user. The data entry system has a predictor trained to predict a next item in the sequence. The predictor comprises a plurality of learnt text item embeddings each text item embedding representing a text item in a numerical form, the text item embeddings having a plurality of different lengths. A projection component obtains text item embeddings of the context text items and projects these to be of the same length. The predictor comprises a trained neural network which is fed the projected text item embeddings and which computes a numerical output associated with the predicted next item.

Abstract: Techniques for question answering involve receiving, from a user, a text input expressing a question in natural language. In response to the question, a text output expressing an answer to the question may be generated. A plurality of documents comprising natural language text may be analyzed, involving mapping the question to one or more hypotheses, analyzing at least one passage of text in at least one of the documents to determine whether the passage entails at least one of the hypotheses, and in response to determining that the passage entails at least one of the hypotheses, identifying the passage as providing supporting evidence for the answer to the question. The answer and the at least one passage identified as providing supporting evidence for the answer may be presented to the user in response to the text input.

Abstract: A system and method for solving general classes of difficult financial calculations such as capital calculations and pricing calculations on a quantum computer are disclosed. In some embodiments, the disclosed method reduces financial problems to problems in # P. In some embodiments, the method includes constructing a quantum circuit whose ground states are comprised of solutions to an associated NP problem that the problem in # P is counting, and then using a quantum annealing process to find one of those ground states, if it exists. For the gross estimate of the size of the solution set of the NP problem, which the # P problem is counting, the method includes finding one solution that achieves a specific random hash value, applying a correction formula, and amplifying the result using a k-fold iterated hash function, in which the number of bits of the hash value increases linearly with each iteration.

Abstract: A method includes identifying costs associated with different outcomes of a failure prediction algorithm. The algorithm is configured to predict one or more faults with at least one piece of industrial equipment. The different outcomes include both successful and unsuccessful predictions by the algorithm. The method also includes identifying a threshold value for the algorithm using the costs, where the threshold value is used by the failure prediction algorithm to identify whether maintenance of the at least one piece of industrial equipment is needed. The method further includes providing the threshold value to the algorithm. The threshold value is selected such that a net positive economic benefit is obtained from use of the threshold value with the failure prediction algorithm. In addition, the method can include generating a signal indicating whether maintenance is needed based on a comparison of an indicator value calculated using the algorithm and the threshold value.

Abstract: Systems and methods for processing loops in computational graphs representing machine learning models are disclosed. An example method begins with obtaining data representing a computational graph. Data identifying an allocation of the computational graph across devices is obtained. Additionally, one or more nodes in the computational graph that represent a respective control flow statement are identified. For each identified node, a structure of nodes and edges that represents an operation that provides a current state of recursion or iteration in the respective control flow statement is generated. This structure is inserted into the computational graph and the allocation of nodes to devices is modified to assign the structure to a device.

Abstract: To identify a scenario that will bear a good simulation result from among a plurality of scenarios used in an agent-based simulation with a reduced amount of computation, there is provided an information processing apparatus comprising a counting part configured to count the number of agents in each of a plurality of states at a middle of a simulation that involves a plurality of agents, and a generation part configured to generate characteristic data used for prediction of a result of the simulation based on the number of agents in each of the plurality of states.

Abstract: A multi-iteration method for compressing a deep neural network into a sparse neural network without degrading the accuracy is disclosed herein. In an example, the method includes determining a respective initial compression ratio for each of a plurality of matrices characterizing the weights between the neurons of the neural network, compressing each of the plurality of matrices based on the respective initial compression ratio, so as to obtain a compressed neural network, and fine-tuning the compressed neural network.

Abstract: A system, method and program product for recommending a network resource provider to a resource consumer. An interactive recommendation engine for determining a recommendation of a network resource provider is provide and includes: a requirements collection manager that collects a set of requirements for an organization and includes: a query management system that provides an interactive platform for implementing a natural language dialog with a user; and a semantic analysis system that analyzes inputs from the user to identify requirements and formulate outputs to the user; a provider data curation manager that curates structured and unstructured provider information into a provider knowledgebase; and a decision analytics system that analyzes the set of requirements and provider knowledgebase to identify a recommended resource provider.