Abstract: Systems and methods for estimating a height of an object from a monocular image are described herein. Objects are detected in the image, each object being indicated by a region of interest. The image is then cropped for each region of interest and the cropped image scaled to a predetermined size. The cropped and scaled image is then input into a convolutional neural network (CNN), the output of which is an estimated height for the object. The height may be represented by a mean of a probability distribution of possible sizes, a standard deviation, as well as a level of confidence. A location of the object may be determined based on the estimated height and region of interest. A ground truth dataset may be generated for training the CNN by simultaneously capturing a LIDAR sequence with a monocular image sequence.

Abstract: The software system processes extracts reliable, significant and relevant patterns. System runs through preprocessing steps. System then generates the size 1 patterns. It then checks for both reliability and refinability of the size 1 patterns. System grows the refinable patterns by increasing the attributes and its values in the pattern by one at a time to find a size 2 pattern. The system then uses the number of pattern occurrences of size 2 pattern as a basis to find the reliable patterns. System also checks for statistical significance over the size 1 patterns and once again for the refinability of the size 2 patterns. System checks for relevance of the size 1 patterns by obtaining the disjointed record complement set. Software system readjusts the pattern statistics of size 1 and removes the non-relevant super-patterns. This process is repeated from size 2 to N.

Abstract: Disclosed is a method and system for detecting outliers in real-time for a univariate time-series signal. The system may receive the univariate time-series signal, comprising a plurality of datasets, from a data source. The system may compute a standard deviation of a dataset of the plurality of datasets. Subsequently, the system may compute the optimal sample block size and the critical sample size of the dataset. Further, the system may determine the optimal operational block size of the dataset. The system may segment the plurality of datasets into blocks based upon the optimal operational block size. The system may detect the outliers by performing an outlier detection technique on the blocks, thereby ensuring improved execution time while minimally affecting precision and accuracy of the outcome of the outlier detection method.

Abstract: A recurrent neural network (RNN) is trained to identify split positions in long content, wherein each split position is a position at which the theme of the long content changes. Each sentence in the long content is converted to a vector that corresponds to the meaning of the sentence. The sentence vectors are used as inputs to the RNN. The high-probability split points determined by the RNN may be combined with contextual cues to determine the actual split point to use. The split points are used to generate thematic segments of the long content. The multiple thematic segments may be presented to a user along with a topic label for each thematic segment. Each topic label may be generated based on the words contained in the corresponding thematic segment.

Abstract: An analytics-based security monitoring system adapted to detect a plurality of behavioral characteristics from behavioral data, each representing an action conducted in a computing environment. Furthermore, the system determines, in accordance with a correlation profile, one or more behavioral fragments, each comprising a plurality of the behavioral characteristics. In accordance with the correlation profile, the one or more determined behavioral fragments are correlated against an attack profile comprising a plurality of sets of behavioral fragments where each set of behavioral fragments forms a malicious behavior pattern of a known attack. Thereafter, an attack based on the correlated one or more determined behavioral fragments may be identified, and the correlation profile is updated after an analysis of the identified attack.

Abstract: A method and a system for automatic image caption generation are provided. The automatic image caption generation method according to an embodiment of the present disclosure includes: extracting a distinctive attribute from example captions of a learning image; training a first neural network for predicting a distinctive attribute from an image, by using a pair of the extracted distinctive attribute and the learning image; inferring a distinctive attribute by inputting the learning image to the trained first neural network; and training a second neural network for generating a caption of an image by using a pair of the inferred distinctive attribute and the learning image. Accordingly, a caption well indicating a feature of a given image is automatically generated, such that an image can be more exactly explained and a difference from other images can be clearly distinguished.

Abstract: In implementations of higher-order network embedding, a computing device maintains interconnected data in the form of a graph that represents a network, the graph including nodes that each represent entities in the network and node associations that each represent edges between the nodes in the graph. The computing device includes a network embedding module that is implemented to determine a frequency of k-vertex motifs for each of the edges in the graph, and derive motif-based matrices from the frequency of each of the k-vertex motifs in the graph. The network embedding module is also implemented to determine a higher-order network embedding for each of the nodes in the graph from each of the motif-based matrices. The network embedding module can then concatenate the higher-order network embeddings into a matrix representation.

Abstract: According to some embodiments, system and methods are provided comprising: receiving data; providing a simulation model for the data; generating one or more simulations via a Bayesian module based on the data, wherein the simulation includes one or more nodes in a chain; executing the Bayesian module to determine the acceptability of the nodes in the simulation based on a Bayesian rule, wherein execution of the Bayesian module further comprises: generating a binary decision tree representing the chain in the simulation, wherein the chain includes one or more nodes; prioritizing which nodes in the tree to simulate; generating one or more simulations; executing the simulation model with data associated with the prioritized nodes in the tree in parallel to determine a posterior probability for each prioritized node; and determining whether each prioritized node is accepted or rejected based on the posterior probabilities. Numerous other aspects are provided.

Abstract: Methods, systems, and computer-accessible mediums are described to authenticate a user using a user's handwriting style rather than using the user's signature through adaptive handwriting challenges which are verified using the machine learning technique of a generative adversarial network.

Abstract: The disclosed computer-implemented method for categorizing security incidents may include (i) generating, within a training dataset, a feature vector for each of a group of security incidents, the feature vector including features that describe the security incidents and the features including categories that were previously assigned to the security incidents as labels to describe the security incidents, (ii) training a supervised machine learning function on the training dataset such that the supervised machine learning function learns how to predict an assignment of future categories to future security incidents, (iii) assigning a category to a new security incident by applying the supervised machine learning function to a new feature vector that describes the new security incident, and (iv) notifying a client of the new security incident and the category assigned to the new security incident. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: Systems and methods are disclosed for end-to-end neural networks for speech recognition and classification and additional machine learning techniques that may be used in conjunction or separately. Some embodiments comprise multiple neural networks, directly connected to each other to form an end-to-end neural network. One embodiment comprises a convolutional network, a first fully-connected network, a recurrent network, a second fully-connected network, and an output network. Some embodiments are related to generating speech transcriptions, and some embodiments relate to classifying speech into a number of classifications.

Abstract: A computer-implemented method may be used for security event monitoring. The method may include receiving data from a first operating system and defining an audit classes data filter for collection by a security event monitoring application. Additionally, the method may include comparing the data with the audit classes data filter and comparing the data with a set of blacklisted values. Additionally, the method may include outputting a common structure format data based on the comparison of the processing data with the audit classes data filter and the blacklisted values.

Abstract: A method for training a deep learning network includes defining a loss function corresponding to the network. Training samples are received and current parameter values are set to initial parameter values. Then, a computing platform is used to perform an optimization method which iteratively minimizes the loss function. Each iteration comprises the following steps. An eigCG solver is applied to determine a descent direction by minimizing a local approximated quadratic model of the loss function with respect to current parameter values and the training dataset. An approximate leftmost eigenvector and eigenvalue is determined while solving the Newton system. The approximate leftmost eigenvector is used as negative curvature direction to prevent the optimization method from converging to saddle points. Curvilinear and adaptive line-searches are used to guide the optimization method to a local minimum. At the end of the iteration, the current parameter values are updated based on the descent direction.

Abstract: Described is a system for pattern recognition designed for neuromorphic hardware. The system generates a spike train of neuron spikes for training patterns with each excitatory neuron in an excitatory layer, where each training pattern belongs to a pattern class. A spiking rate distribution of excitatory neurons is generated for each pattern class. Each spiking rate distribution of excitatory neurons is normalized, and a class template is generated for each pattern class from the normalized spiking rate distributions. An unlabeled input pattern is classified using the class templates. A mechanical component of an autonomous device can be controlled based on classification of the unlabeled input pattern.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for depth concatenation using a matrix computation unit. One of the methods includes: receiving a request to process network inputs to a neural network using an integrated circuit, the neural network comprising a depth concatenation neural network layer; and generating instructions that, when executed by the integrated circuit, cause the integrated circuit to perform operations comprising: for each spatial location in a first input tensor to the depth concatenation layer and a second input tensor to the depth concatenation layer: multiplying, using the matrix computation unit, a second depth vector for the spatial location by a shift weight matrix for the depth concatenation layer to generate a shifted second depth vector; and adding the shifted second depth vector and a first input depth vector for the spatial location to generate a concatenated depth vector.

Abstract: Disclosed are a method, a device, a system and/or a manufacture of data analysis and prediction of a dataset through algorithm extrapolation from a spreadsheet formula. In one embodiment, a method extracts one or more spreadsheet formulas from one or more cells of a spreadsheet file and assembles a formula algorithm. The formula algorithm accepts a set of data entries comprising one or more independent variables, and outputs a prediction metric as a dependent variable such that each data entry is calculation independent. An extrapolated algorithm expressed in a programming language is generated. A computation block of a dataset is specified and submitted for computation along with the extrapolated algorithm over a network. The dataset comprising two or more data entries usable as an input to the extrapolated algorithm. An output data re-combined from the first output block and one or more additional output blocks is received.

Abstract: An advisor creates configurations for generating multi-representations of time series data based upon detected characteristics such as length, interval, minimums, data types, etc., as well as configurations provided by a user. In an offline mode the advisor may further consider a previous time series workload. In an on-line mode the advisor may adapt multi-representation configurations with respect to ongoing changes in a current time series workload. The advisor may reference a cost model including values quantifying various dimensions (e.g., compression technique, accuracy, covered time period, storage medium, memory consumption, speed) of the multi-representations for optimization purposes. Configurations created by the advisor may be input to a storage engine to generate and store the multi-representations according to goals for data aging, operation execution pattern optimization, and ease of access to time series data located in hot zones.

Abstract: A computer-implemented method of treating a patient's and automated enteral feeding, comprising: monitoring a plurality of reflux-related parameters and at least one reflux event while the patient is automatically enterally fed by an enteral feeding controller according to a baseline feeding profile including a target nutritional goal, training a classifier component of a model for predicting likelihood of a future reflux event according to an input of scheduled and/or predicted plurality of reflux-related parameters, the classifier trained according to computed correlations between the plurality of reflux-related parameters and the at least one reflux event, feeding scheduled and/or predicted reflux-related parameters into the trained classifier component of the model for outputting risk of likelihood of a future reflux event, and computing, by the model, an adjustment to the baseline feeding profile for reducing likelihood of the future reflux event and for meeting the target nutritional goal.

Abstract: In some implementations there may be provided a system. The system may include a processor and a memory. The memory may include program code which causes operations when executed by the processor. The operations may include analyzing a series of events contained in received data. The series of events may include events that occur during the execution of a data object. The series of events may be analyzed to at least extract, from the series of events, subsequences of events. A machine learning model may determine a classification for the received data. The machine learning model may classify the received data based at least on whether the subsequences of events are malicious. The classification indicative of whether the received data is malicious may be provided. Related methods and articles of manufacture, including computer program products, are also disclosed.

Abstract: Methods and systems for optimising the quality of visual data. Specifically, methods and systems for preserving visual information during compression and decompression. An example method for optimising visual data includes using a pre-processing hierarchical algorithm to optimise visual data prior to encoding the visual data in visual data processing; and using a post-processing hierarchical algorithm to enhance visual data following decoding visual data in visual data processing.

Abstract: The disclosed computer-implemented method may include identifying data format requirements for one or more machine-learning-based audio/video classifiers. The classifiers may be configured to detect classifiable features of decoded audio/video data. The method may also include decoding once, for the one or more classifiers, a video stream into audio/video data based on the identified data format requirements, and creating a new instance of each of the one or more classifiers. In addition, the method may include transforming the audio/video data for each instance based on the respective data format requirements and providing the respective transformed audio/video data to each instance. The method may also include performing, in parallel, classification of each transformed audio/video data by each respective instance and then terminating each instance of the one or more classifiers. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: A system and method for outputting modified input data for storage comprises a communication interface, a comparison module, a translation module and an output module. The communication interface is arranged to receive an input data set comprising a plurality of data labels. The comparison module is arranged to compare the data labels to a plurality of nomenclature-labels in a nomenclature database and identify an undefined data label by determining that at least one of the data labels is not present in the nomenclature database, based on the comparison. The translation module is arranged to translate the undefined data label into a nomenclature-label using a synonyms database. The output module is arranged to output a modified data set based on the input data set and the translated undefined label for storage.

Abstract: Disclosed is method for training a plurality of visual processing algorithms for processing visual data. The method includes using a pre-processing hierarchical algorithm to process the visual data prior to encoding the visual data in visual data processing, and using a post-processing hierarchical algorithm to further process the visual data following decoding visual data in visual data processing. The encoding and decoding are performed with respect to a predetermined visual data codec and may be content specific.

Abstract: Embodiments of a system for determining personal attributes based on public interaction data are illustrated. In one embodiment, the system employs a process for predicting personal attributes based on public interaction data by constructing matrices based on user interactions drawn from public posts on a social media website. The process may further learn a compact representation for a plurality of users based on public posts using the matrices, extract the compact representation of one or more users that have been labeled, and apply a classifier to learn about a particular personal attribute. Through this, a prediction of personal attributes of users that have not been labeled may be obtained.

Abstract: An abnormal detection apparatus including an imaging unit configured to image generate a first and second image frames included in a first image frame group; a pseudo work generation unit configured to generate the first and a third image frames, the third and the second image frames, or the third and a fourth image frames, included in a second image frame group, respectively, with respect to the first and second image frames included in the first image frame group; a normal space generation unit configured to generate a normal space data based on the first and second image frames included in the first image frame group, and the first and third image frames, the third and second image frames, or the third and fourth image frames, included in the second image frame group; and a comparison unit configured to detect abnormality based on the normal space data.

Abstract: A system and method for botmaster discovery are disclosed. The system and method may be used in a network that has a plurality of known malicious domains, a plurality of servers each having a known malicious internet protocol (IP) address in which each server is associated with one or more of the plurality of domains, a plurality of hosts associated with one or more of the plurality of servers wherein the host is one of a bot which is compromised host and involved as a part of resource for cyber-crime purpose and a botmaster which involves bots for cyber-crime purpose.

Abstract: A system and method of identifying suspicious item-related features are disclosed. In some embodiments, a sample of item listings is received. Each item listing in the sample of item listings may correspond to an item offered for sale on an e-commerce website. Item-related data for each item listing in the sample of item listings may be extracted. Training data comprising identifications of which item listings in the sample of item listings are suspicious and identifications of which item listings in the sample of item listings are not suspicious may be received. A model for classifying a candidate item listing on the e-commerce website as being suspicious may be determined using the training data.

Abstract: Methods, systems, and computer programs are presented for training, with adversarial constraints, a student model for speech recognition based on a teacher model. One method includes operations for training a teacher model based on teacher speech data, initializing a student model with parameters obtained from the teacher model, and training the student model with adversarial teacher-student learning based on the teacher speech data and student speech data. Training the student model with adversarial teacher-student learning further includes minimizing a teacher-student loss that measures a divergence of outputs between the teacher model and the student model; minimizing a classifier condition loss with respect to parameters of a condition classifier; and maximizing the classifier condition loss with respect to parameters of a feature extractor. The classifier condition loss measures errors caused by acoustic condition classification. Further, speech is recognized with the trained student model.

Abstract: An acceleration apparatus applied in an artificial neuron is disclosed. The acceleration apparatus comprises an AND gate array, a first storage device, a second storage device and a multiply-accumulate (MAC) circuit. The AND gate array with plural AND gates receives a first bitmap and a second bitmap to generate an output bitmap. The first storage device stores a first payload and outputs a corresponding non-zero first element according to a first access address associated with a result of comparing the first bitmap with the output bitmap. The second storage device stores a second payload and outputs a corresponding non-zero second element according to a second access address associated with a result of comparing the second bitmap with the output bitmap. The MAC circuit calculates a dot product of two element sequences from the first storage device and the second storage device.

Abstract: Systems and methods for determining trust levels for components of a computing application using a blockchain. The system may include a development framework, a trust matrix, a trust level calculation module, a visual design subsystem, and a deployment subsystem, where trust levels are associated with components, combinations of components, graphs, and blueprints, where trust levels relate to categories of use.

Abstract: This application discloses a method implemented by an electronic device to detect a signature event (e.g., a baby cry event) associated with an audio feature (e.g., baby sound). The electronic device obtains a classifier model from a remote server. The classifier model is determined according to predetermined capabilities of the electronic device and ambient sound characteristics of the electronic device, and distinguishes the audio feature from a plurality of alternative features and ambient noises. When the electronic device obtains audio data, it splits the audio data to a plurality of sound components each associated with a respective frequency or frequency band and including a series of time windows. The electronic device further extracts a feature vector from the sound components, classifies the extracted feature vector to obtain a probability value according to the classifier model, and detects the signature event based on the probability value.

Abstract: The invention utilizes a two-component system to detect third party security threats and drive internal system processes based on the detection. The first component of the system is a threat level engine, which collects external and internal system data on a real-time basis to determine changes in conditions that may give rise to a security threat. Based on the external and internal data, the level engine may calculate a threat assessment level to determine the level of the threat. The second component of the system is a third party analytics engine, which may comprise a machine learning component which is configured to detect threat patterns and anomalies, which may in turn be used to trigger events or to drive internal system processes.

Abstract: Systems and methods to code medical records using weighted belief networks are provided. Medical records are received, and may be subjected to pre-processing which includes deduplication of records, indexing the records, meta-tagging the records, and annotating the records. An entity extractor then generates entity dictionaries from public sources. A network creator generates a belief network based on medical relationships. An annotation aligner receives normalized annotations of historical medical records, and a network weighter assigns probability values to the belief network using the normalized annotations to generated a weighted belief network. A health care code classifier utilizes the weighted belief network to classify the medical records by comparing entities within the medical records.

Abstract: Techniques are generally described for performing outfit recommendation using a recurrent neural network. In various examples, a computing device may receive a first state vector representing an outfit comprising at least one article of clothing. First image data depicting a second article of clothing of a first clothing category may be received. A recurrent neural network may generate a first output feature vector based on the first state vector, the first image data and the first clothing category. The first output feature vector may be compared to other feature vectors representing other articles of clothing in the first category to determine distances between the first output feature vector and the other feature vectors. A set of articles of clothing may be recommended based on the distances between the first output feature vector and the other feature vectors.

Abstract: A technique for responding to user input includes assigning respective initial confidence scores to relationship n-tuples in a knowledge graph (KG). Each of the n-tuples designates at least a first entity, a second entity, and a relationship between the first and second entities or a single entity and a relationship between the single entity and one or more properties of the single entity. Respective feature vectors are associated with each of the n-tuples. A training set that includes at least a subset of the n-tuples labeled with respective ground truth labels is generated. Respective initial weights are learned for the feature vectors based on the training set. Respective subsequent confidence scores are generated for each of the n-tuples based on the initial weights for the feature vectors. A response to user input is generated based on the subsequent confidence scores for one or more of the n-tuples.

Abstract: A technique for assigning confidence scores to relationship entries in a knowledge graph includes assigning respective initial confidence scores to relationship n-tuples in a knowledge graph. Each of the relationship n-tuples designates at least a first entity, a second entity, and a relationship between the first and second entities or a single entity and a relationship between the single entity and one or more properties of the single entity. Respective feature vectors are associated with each of the relationship n-tuples. A training set that includes at least a subset of the relationship n-tuples labeled with respective ground truth labels is generated. Respective initial weights are learned for the feature vectors based on the training set. Respective subsequent confidence scores are generated for each of the relationship n-tuples based on the initial weights for the feature vectors.

Abstract: In embodiments of statistics time chart interface cell mode drill down, a first interface displays in a table format that includes columns each having a column heading comprising a different value, each different value associated with a particular event field, and includes one or more rows, each row having a time increment and aggregated metrics that each represent a number of events having a field-value pair that matches the different value represented in one of the columns and within the time increment over which the aggregated metric is calculated. A cell can be emphasized that includes one of the aggregated metrics in a row that includes the respective time increment, and in response, a menu displays options to transition to a second interface.

Abstract: A system for automatically discovering services operating on a network including a service discovery database configured to store expected service behavioral characteristics and service identities of the services operating on the network, a set of service discovery modules configured to collect service behavioral data of the services operating on the network, and a service discovery module controller communicatively coupled to the service discovery module database and the set of service discovery modules, the service discovery module controller configured to generate service behavioral characteristics from the service behavioral data, analyze the service behavioral characteristics using the expected service behavioral characteristics, resulting in a first behavioral analysis, identify a first service identity of at least one service operating on the network from the first behavioral analysis and an association of the first service identity and the expected service behavioral characteristics.

Abstract: Systems and methods may include receiving performance data of components in a system. The performance data may include data for parameters for each of the components. The systems and methods may include determining aggregate data for each group of similar components of the components. The aggregate data for each group of similar components may include a group characteristic for each of the parameters. The systems and methods may include, for each group of similar components, determining whether the data for each of the parameters for each component is consistent with the group characteristic for the respective parameter. The systems and methods may include, for each component of the respective group determining that the component is anomalous in response to determining that the data for a parameter for the component is not consistent with the group characteristic for the parameter.

Abstract: Provided are a neuromorphic computing device, memory device, system, and method to maintain a spike history for neurons in a spiking neural network. A neural network spike history is generated in a memory device having an array of rows and columns of memory cells. There is one row of the rows for each of a plurality of neurons and columns for each of a plurality of time slots. Indication is made in a current column in the row of the memory cells for a firing neuron that a spike was fired. Indication is made in the current column in rows of memory cells of idle neurons that did not fire that a spike was not fired. Information in the array is used to determine a timing difference between a connected neuron and the firing neuron and to adjust a weight of the connecting synapse.

Abstract: Enhancements to network security are provided by identifying malicious actions taken against servers in a network environment, without having to access log data from individual servers. Seed data are collected by an administrator of the network environment, from honeypots and servers whose logs are shared with the administrator, to identify patterns of malicious actions to access the network environment. These patterns of use include ratios of TCP flags in communication sessions, entropy in the use of TCP flags over the life of a communication session, and packet size metrics, which are used to develop a model of characteristic communications for an attack. These attack models are shared with servers in the network environment to detect attacks without having to examine the traffic logs of those servers.

Abstract: A method for identifying a level of similarity between binary vectors includes storing, by a processor on a computing device, in each of a plurality of memory cells on the computing device, one of a plurality of binary vectors, each of the plurality of memory cells including a bitwise comparison circuit. The processor provides, to each of the plurality of memory cells, a received binary vector. Each of the bitwise comparison circuits determines a level of overlap between the received binary vector and the binary vector stored in the memory cell associated with the bitwise comparison circuit. Each of the comparison circuits that determines that the level of overlap satisfies a threshold provides, to the processor, an identification of the stored binary vector with the satisfactory level of overlap. The processor provides an identification of each stored binary vector satisfying the threshold.

Abstract: Examples of the present disclosure improve decision-making for input understanding to assist in determining how to best respond to a user input. A received input is analyzed using an input recognition component, input understanding component and input context component. Potential response options are determined. If uncertainty exists with respect to responding to the received input, an uncertainty value and a cost of misclassification are generated for the potential response options to assist in making a decision as to how to best respond to the received input. The uncertainty value is determined for a potential response and parameters associated with the potential response and the cost of misclassification is a cost associated with pursuing a potential response if the potential response turns out to be incorrect. A response is selected to transmit to a user based on analyzing the generated uncertainty value and the generated cost of misclassification for the potential responses.

Abstract: In embodiments of statistics time chart interface cell mode drill down, a first interface displays in a table format that includes columns each having a column heading comprising a different value, each different value associated with a particular event field, and includes one or more rows, each row having a time increment and aggregated metrics that each represent a number of events having a field-value pair that matches the different value represented in one of the columns and within the time increment over which the aggregated metric is calculated. A cell can be emphasized that includes one of the aggregated metrics in a row that includes the respective time increment, and in response, a menu displays options to transition to a second interface.

Abstract: A coflow identification method includes: obtaining a weighted matrix by means of learning according to historical data in the network, where the weighted matrix is used to minimize a feature distance between data streams belonging to a same coflow and maximize a feature distance between data streams belonging to different coflows; computing a feature distance between any two data streams in the network according to metrics in the data stream layer data feature, the application layer data stream feature distance, the terminal aspect data feature distance, and the weighted matrix; and dividing the data streams in the network into several cluster sets by using a clustering algorithm and according to the feature distance between the any two data streams in the network, where each of the several cluster sets is a coflow.

Abstract: The present disclosure describes methods, systems, and computer program products for performing integration logic programming. One computer-implemented method includes receiving, by operation of a middleware system, first information in a first format corresponding to a first application, applying, by operation of the middleware system, one or more integration logic programming (ILP) patterns to the first information, the one or more ILP patterns representing application integration semantics using a logic programming language, generating, by operation of the middleware system, a second information in response to applying the one or more ILP patterns to the first information, and outputting, by operation of the middleware system, the second information in a second format corresponding to a second application.

Abstract: A system and method for improved categorization and sentiment analysis which is fed textual data such as transcriptions or collated data from a network enabled service, or some other source, which then segments textual data into chunks, parses the data chunks, and analyzes it using a plurality of techniques and metadata gathering methods to determine the sentiment of participating individuals concerning entities mentioned in the textual data and to categorize the discussions, for the purpose of taking actions to improve business outcomes.

Abstract: A neural network is structured to connect the input values of an input set, at each level, to that level's output using a linear bypass connection. The linear bypass connection passes the input values, to the output, without applying a non-linear function to them.

Abstract: A method for a data placement that attempts to predict the most suitable placement, in terms of data reduction, of a newly created storage volume based on the volumes known attributes and the current placement of volumes to deduplication domains is disclosed. The system uses machine learning to perform improved deduplication-aware placement. The system attempts to predict the deduplication domain where a newly created volume would eventually have the best content sharing. The system does this by using the known attributes of the volume at the time of creation, such as owner, volume name, initial size, creation time, and the history of data already in the system and its placement.

Abstract: A flexible approach for storing time series data, utilizes multiple representations in order to achieve optimization amongst various dimensions such as covered time period, accuracy, compression model, and storage medium. A first aspect selectively provides different representations for various portions of a time series along the time axis (horizontal). In this horizontal aspect, a first compression model may be applied to store data occurring during one period in a time series, while a different compression model is applied to store data during another period. In a separate (vertical) aspect, time series data over a same time period may be saved in multiple representations using different compression models. While increasing demands on memory, such vertical storage can afford efficient access for specific purposes (i.e., analytical).