Abstract: A method of utilizing a computing device to correct source data used in machine learning includes receiving, by the computing device, first data. The computing device corrects the source data via an application of a covariate shift to the source data based upon the first data where the covariate shift re-weighs the source data.

Abstract: Techniques are provided for identifying interactions between natural medicines and modern medicines. A corpus of information is analyzed pertaining to interactions of natural medicines with modern medicines for treating a condition. A mapping is created, wherein natural medicines are mapped to modern medicines for treating the condition. An on-line clinical trial study group is created, allowing participation by users, wherein a participation in a group is allowed for a selected set of users based on a group medication usage profile. Information from participants in each group is collected, and the collected information is analyzed.

Abstract: A system includes a memory having instructions therein and at least one processor configured to execute the instructions to: receive a natural language question; determine, from a chat log comprising a plurality of chat session logs, a set of chat session logs most relevant to the natural language question; determine a respective plurality of non-overlapping text spans most relevant to the natural language question within each of a respective plurality of conceptual pseudo-documents; determine a conceptual pseudo-document most relevant to the natural language question; extract a question-answer pair most relevant to the natural language question from the most relevant pseudo-document; and convey the most relevant question-answer pair to a user. Each one of the conceptual pseudo-documents corresponds to a respective one of the most relevant chat session logs.

Abstract: Provided is a system for removing an electronic component from a printed circuit board (PCB). The system may comprise a heating well configured to hold a rework liquid. The system may further comprise a head system configured to create a liquid-tight seal around an electronic component. The system may further comprise a nozzle and a mechanical capture device disposed within the head system. The mechanical capture device may be configured to attach to the electronic component. The system may further comprise a controller. The controller may be configured to release the rework liquid through the nozzle and onto the electronic component and lift the electronic component off the PCB.

Abstract: A computer-implemented method includes determining that an update to data on a local computer has been initiated by an application and sending a notification to a remote computer that the update has been initiated. The method includes sending, to the remote computer, an indication of the data on the local computer being updated. The method includes, in response to determining that updates to the data on the local computer and meta data corresponding to the data have been completed, sending a second notification to the remote computer that the update has been completed.

Abstract: Aspects of the present invention disclose a method for handling incoming microservice requests at an application server. The method includes one or more processors identifying a microservice request of a user at an application server. The method further includes querying a database associated with the application server for a feedback history corresponding to the microservice request. The method further includes collecting a response corresponding to the microservice request of the user. The method further includes generating a rule corresponding to the microservice request based at least in part on the response corresponding to the microservice request.

Abstract: A computer-implemented method identifies one or more concepts from a document, wherein the document comprises information on completing a task. The method computes a relevance value of an utterance with respect to completing the task using the one or more identified concepts. The method removes the utterance from a dialog model to be used for completing the task when the relevance value of the utterance is below a given threshold value.

Abstract: Aspects include configuring a plurality of functional self-test controllers in a test control device to run a plurality of functional test cases in parallel on a plurality of devices under test. Test traffic is arbitrated between the functional self-test controllers and a plurality of packeted protocol layer interfaces of the test control device. One or more protocol specific conversions are performed between the test traffic and a device-specific packeted protocol of each of the devices under test. Payload checking is performed between the packeted protocol layer interfaces and the devices under test to verify responses of the devices under test to the functional test cases.

Abstract: A method for reducing incidents of data loss in redundant arrays of independent disks (RAIDs) having the same RAID level is disclosed. In one embodiment, such a method identifies, in a data storage environment, a set of RAIDs having a common RAID level. The method also identifies, in the set of RAIDs, higher risk storage drives having a failure risk above a threshold and lower risk storage drives having a failure risk below the threshold. The method swaps, within the RAIDs, higher risk storage drives with lower risk storage drives to more evenly distribute higher risk storage drives across the RAIDs. A corresponding system and computer program product are also disclosed.

Abstract: Embodiments of the present disclosure relate to a method, system, and computer program product for exception handling. In some embodiments, a method is disclosed. According to the method, in response to an application throwing an exception, a target stack frame for handling the exception is detected from a call stack of the application. The call stack comprises a plurality of stack frames. One of the plurality of stack frames corresponds to a function called by the application and supports either a one-phase exception handling mechanism or a two-phase exception handling mechanism. In response to the target stack frame being detected from the call stack, the target stack frame is caused to handle the exception. In other embodiments, a system and a computer program product are disclosed.

Abstract: Methods, systems and computer program products for query processing are provided herein. A computer-implemented method includes receiving a first query from a user, determining whether the first query is capable of being answered using symbolic reasoning performed on data of a symbolic knowledge base, and executing the symbolic reasoning to generate a first query answer in response to a determination that the first query is capable of being answered using the symbolic reasoning. Axioms are extracted from a plurality of documents when it is determined that a second query is not capable of being answered using the symbolic reasoning. The method further includes determining whether the axioms are consistent with the symbolic knowledge base, and generating a second query answer based on the axioms in response to a determination that the one or more axioms are consistent with the symbolic knowledge base.

Abstract: A computer-implemented method for advancing speculative execution in microarchitectures is disclosed. A non-limiting example of the computer-implemented method includes receiving, by a processor, a test scenario including a first load instruction from a first memory location flagged with a delay notification and a speculative memory access instruction from a second memory following the first load instruction. The method executes, by the processor, the first load instruction from the first memory location and delays a return of data from the first memory location for a number of processor cycles. The method executes, by the processor, the speculative storage access instruction from the second memory location during the delay in returning the data from the first memory location.

Abstract: An aspect of the invention includes receiving machine learning (ML) training data that includes a plurality of features for a plurality of observations. The ML training data is broken into a plurality of non-overlapping subsets of features and observations. A first ML algorithm is trained based on a first subset of the features and observations, and a second ML algorithm is trained based on a second subset of the features and observations. The training of the first ML algorithm overlaps in time with the training of the second ML algorithm. The first and second ML algorithms are tested. Either the first or second ML algorithm is selected based at least in part on results of the testing. The selected ML algorithm is retained as a trained ML algorithm for predicting one or more of the plurality of features based on one or more others of the plurality of features.

Abstract: A method of forming vertical fins on a substrate at the same time, the method including, forming a mask segment on a first region of the substrate while exposing the surface of a second region of the substrate, removing a portion of the substrate in the second region to form a recess, forming a fin layer in the recess, where the fin layer has a different material composition than the substrate, and forming at least one vertical fin on the first region of the substrate and at least one vertical fin on the second region of the substrate, where the vertical fin on the second region of the substrate includes a fin layer pillar formed from the fin layer and a substrate pillar.

Abstract: A device and method for fabricating a photovoltaic device includes forming a double layer transparent conductive oxide on a transparent substrate. The double layer transparent conductive oxide includes forming a doped electrode layer on the substrate, and forming a buffer layer on the doped electrode layer. The buffer layer includes an undoped or p-type doped intrinsic form of a same material as the doped electrode layer. A light-absorbing semiconductor structure includes a p-type semiconductor layer on the buffer layer, an intrinsic layer and an n-type semiconductor layer.

Abstract: A method is provided for validating a hypothesis sentence. The method extracts, from a document database D using a hypothesis sentence that includes a causal part and a consequence part, a set D1 of documents related to the causal part. The method extracts, from set D1, a set S of sentences that include expressions of opinion. The method obtains a word list W of words that have a high co-occurrence in the set D1. The method selects, from the set S, a set S1 of sentences that are positionally close to any of the words in the word list W. The method selects, from the set S, a set S2 of sentences that are related to the words in the consequence part of the hypothesis. The method extracts and displays sentences included in both the set S1 and the set S2 as opinion sentences in relation to the hypothesis sentence.

Abstract: A method, system, and computer program product for detecting and monitoring a signal is provided. The method includes detecting an alignment point for a periodic signal segment of a periodic signal generated by an apparatus being monitored for standard functionality In response, the apparatus is activated from a period prior to the alignment point to an end point of the periodic signal segment and a first point of the periodic signal segment is located. Likewise, a second point of an additional periodic signal segment of the periodic signal is located. The periodic signal is normalized based on results of locating the first point and the second point.

Abstract: Shared memory communication is facilitated between systems of a computing environment capable of communicating over a network using transmission control protocol/Internet protocol (TCP/IP). The network includes a network path between one system and another system of the computing environment, where the network path passes through one or more routers of the network. The facilitating includes obtaining performance-related data for shared memory communication of the one system with the other system across the network path using a remote direct memory access (RDMA) protocol. Based on the performance-related data, the facilitating includes dynamically determining whether to use the RDMA protocol for shared memory communication of the one system with the other system across the network path, rather than the TCP/IP protocol.

Abstract: According to one embodiment, a method, computer system, and computer program product for navigating driverless vehicles is provided. The present invention may include ingesting data pertaining to the operation of the driverless vehicle, utilizing that data to predict tasks, which are driverless vehicle service tasks such as parking, maintenance, fueling, et cetera. The invention may further include determining the risk that a user may have need of the driverless vehicle, and scheduling the tasks to provide a balanced combination of convenience to the user, effective maintenance of the driverless vehicle, cost, and time. The method further includes navigating the driverless vehicle to accomplish the scheduled tasks.

Abstract: A semiconductor device includes a field-effect transistor, a first back-end-of-line (BEOL) metallization level and a second BEOL metallization level disposed above the first BEOL metallization level. A portion of the field-effect transistor includes lithium therein, and the field-effect transistor is integrated between the first and second BEOL metallization levels. The portion of the field-effect transistor including the lithium therein can be a channel layer, or a source and/or drain region.

Abstract: An illustrative embodiment includes a method for protecting a machine learning model. The method includes: determining concept-level interpretability of respective units within the model; determining sensitivity of the respective units within the model to an adversarial attack; identifying units within the model which are both interpretable and sensitive to the adversarial attack; and enhancing defense against the adversarial attack by masking at least a portion of the units identified as both interpretable and sensitive to the adversarial attack.

Abstract: A plurality of files are stored on magnetic tape having at least one wrap for storing data sets. Each data set includes one or more records. A number of data sets that will remain in a buffer memory after performing a read ahead operation is determined. A number of records that are likely to remain in the buffer memory after performing a read ahead operation is determined. For each record likely to remain in the buffer memory after performing a read ahead operation, positions of the records on the magnetic tape are determined. Each file of the plurality of files are classified into one group of a plurality of groups based on a position of a record corresponding with each respective file and the number of records that are likely to remain in the buffer memory after performing a read ahead operation.

Abstract: A computer-implemented method, system and computer program product for improving prediction accuracy in machine learning techniques. A teacher model is constructed, where the teacher model generates a weight for each data case. The current student model is then trained using training data and the weights generated by the teacher model. After training the current student model, the current student model generates state features, which are used by the teacher model to generate new weights. A candidate student model is then trained using training data and these new weights. A reward is generated by comparing the current student model with the candidate student model using training and testing data, which is used to update the teacher model if a stopping rule has not been satisfied. Upon a stopping rule being satisfied, the weights generated by the teacher model are deemed to be the “optimal” weights which are returned to the user.

Abstract: Mechanisms are provided to implement an enhanced privacy deep learning system framework (hereafter “framework”). The framework receives, from a client computing device, an encrypted first subnet model of a neural network, where the first subnet model is one partition of multiple partitions of the neural network. The framework loads the encrypted first subnet model into a trusted execution environment (TEE) of the framework, decrypts the first subnet model, within the TEE, and executes the first subnet model within the TEE. The framework receives encrypted input data from the client computing device, loads the encrypted input data into the TEE, decrypts the input data, and processes the input data in the TEE using the first subnet model executing within the TEE.

Abstract: The present invention provides a method and system for analyzing human speech during natural language processing interactions between humans and computers to aid in computer learning. The method processes human language tutorial videos each having a visual track, an audio track and captions. Multiple videos are simultaneously processed in parallel using stream processing to identify spoken words or phrases in the videos by comparing them with benchmark words/phrases stored on a computer. Confidence scores are determined for each of the spoken words/phrases which are assigned to a list of the benchmark words/phrases on the computer when a threshold value is met. A system administrator can identify spoken words/phrases to which the threshold value is not met.

Abstract: Embodiments for efficient machine and deep learning hyperparameter tuning in a distributed computing system. Runtime metrics of each training iteration are collected to identify candidate jobs to merge during an execution phase. The candidate jobs are grouped into job groups, and the job groups containing the candidate jobs are merged together subsequent to each iteration boundary for execution during the execution phase.

Abstract: Distributing machine learning models to electronic devices is provided. A number of machine learning models are built and stored on a server, wherein each model is customized to different device and user requirements. A request for a machine learning model is received from an electronic device, wherein the request includes a device profile comprising device attributes and user characteristics. The server determines which machine learning model from among the number of models is best suited for the device profile according to a set of model policies. The machine learning model best suited for the device profile is selected and downloaded to the device.

Abstract: A shipping box includes a plurality of sides and a plurality of bladders. The sides are arranged as a rectangular solid with fixed sides and a top side. The top side includes an openable cover for access to an interior of the rectangular solid. The rectangular solid has a size determined based on dimensions of an object placed in the interior of the rectangular solid. The bladders are arranged on each of the sides and configured to be inflated as a cushioning mechanism for the object while placed in the interior of the rectangular solid. The bladders are made of a material based on a weight of the at least one object.

Abstract: A system configuration template is configured according to a first configuration command received from a configuration management system executing on a control computer system. A node configuration command is generated for a node computer system. The node configuration command adjusts a configuration of a node computer system to conform to the system configuration template.

Abstract: A quantum computing device is formed using a first chip and a second chip, the first chip having a first substrate, a first set of pads, and a set of Josephson junctions disposed on the first substrate. The second chip has a second substrate, a second set of pads disposed on the second substrate opposite the first set of pads, and a second layer formed on a subset of the second set of pads. The second layer is configured to bond the first chip and the second chip. The subset of the second set of pads corresponds to a subset of the set of Josephson junctions selected to avoid frequency collision between qubits in a set of qubits. A qubit is formed using a Josephson junction from the subset of Josephson junctions and another Josephson junction not in the subset being rendered unusable for forming qubits.

Abstract: A computer-implemented method and system for resilient adaptive biased locking. The method includes adding, in a system including an adaptive lock reservation scheme having a learning state, a component comprising a per class counter that counts, collectively, a number of learning failures and a number of revocation failures. An embodiment includes initializing the per class counter upon loading a class with a predetermined value representing at least one of a maximum number of learning failures and cancellation instances associated with the class. An embodiment includes initializing, based on a determination of an operational state of the per class counter for an object transitioning from one of the learning state and a biased state to a flatlock state, a lock word of the object directly to the flatlock state while bypassing the biased state.

Abstract: A lesion detection and classification artificial intelligence (AI) pipeline comprising a plurality of trained machine learning (ML) computer models is provided. First ML model(s) process an input volume of medical images (VOI) to determine whether VOI depicts a predetermined amount of an anatomical structure. The AI pipeline determines whether criteria, such as a predetermined amount of an anatomical structure of interest being depicted in the input volume, are satisfied by output of the first ML model(s). If so, lesion processing operations are performed including: second ML modal(s) processing the VOI to detect lesions which correspond to the anatomical structure of interest; third ML model(s) performing lesion segmentation and combining of lesion contours associated with a same lesion; and fourth ML models processing the listing of lesions to classify the lesions. The AI pipeline outputs the listing of lesions and the classifications for downstream computing system processing.

Abstract: A computer transforms content of a composite table into structured data objects. The computer receives a composite table and identifying a data zone characterized by data columns, and a header zone. The computer identifies first header cells arranged coextensive with a single data column and second header cells arranged coextensive with a set of data columns. The computer generates a hierarchical representation of said header cells, based at least in part, on the header cell arrangements. The computer generates a revised table based on the hierarchical representation, with the first header cells identifying a data column and the second header cells identify a first header cell. The computer generates structured data objects representing the zones and being arranged based, at least in part, on the revised table and where the structured data objects are keyed to the first header cells.

Abstract: In an approach to summarizing contextually sensitive text documents based on long short-term memory networks, one or more computer processors create a model based on a vectorized document corpus. The one or more computer processors determine one or more contextually relevant sections utilizing the created model. The one or more computer processors generate a relevancy level for each of the one or more respective determined sections. The one or more computer processors prioritize each of the one or more determined sections based on the respective relevancy level. The one or more computer processors generate one or more textual summaries based on the prioritized sections, wherein the generated one or more textual summaries are aggregated into a document.

Abstract: Embodiments for intelligent monitoring of a health state of a user by a processor. A health state of a user may be learned while engaged in one or more activities associated with a computing device. One or more mitigating actions may be identified and recommended to implement by the user to minimize one or more possible negative impacts upon the health state of the user while engaged in the one or more activities associated with the computing device.

Abstract: Methods, computer program products, and systems are presented. The method computer program products, and systems can include, for instance: obtaining from a user one or more data queries; identifying a product of interest in response to the one or more data query; examining a plurality of product records to determine a set of related products that are related to the product of interest, wherein the examining includes performing image analysis to extract one or more product topic classifier from product image data representing one or more product; and providing one or more output in response to the examining.

Abstract: A method includes determining, based on an input data sample, a set of probabilities. Each probability of the set of probabilities is associated with a respective label of a set of labels. A particular probability associated with a particular label indicates an estimated likelihood that the input data sample is associated with the particular label. The method includes modifying the set of probabilities based on a set of adjustment factors to generate a modified set of probabilities. The set of adjustment factors is based on a first relative frequency distribution and a second relative frequency distribution. The first relative frequency distribution indicates for each label of the set of labels, a frequency of occurrence of the label among training data. The second relative frequency distribution indicates for each label of the set of labels, a frequency of occurrence of the label among post-training data provided to the trained classifier.

Abstract: In an approach to improve virtual reality (VR) by rendering objects in a viewing area within a virtual environment as static embodiments display, by a user interface on a computing device, a plurality of captured frames of a physical surrounding area. Additionally, embodiments determine a primary focus area and a secondary focus area of the displayed plurality of captured frames using object recognition techniques and detect movement of a second object in the secondary focus area. Additionally, responsive to detecting movement of the second object in the secondary focus area, embodiments render the second object as static in the displayed plurality of captured frames.

Abstract: A memory array with memory cells may have one or more heaters integrated into the memory array between the memory cells. A processor in communication with the heater may notify the heater to activate when a trigger event occurs.

Abstract: The present disclosure provides for medical device administration and interaction by identifying a medical event affecting a first individual, a medical device associated with treating the medical event, and an initial attentiveness level of an administering individual; outputting a tutorial for how to use the medical device to treat the medical event with a first control level, wherein the first control level based on the initial attentiveness level of the administering individual; and in response to receiving a triggering event while outputting the tutorial: determining a current attentiveness level for the administering individual; and in response to the current attentiveness level indicating a change from the initial attentiveness level, adjusting the tutorial based on a second control level, different than the first control level, based on the change from the initial attentiveness level.

Abstract: Techniques for outside-in mapping for corpus pairs are provided. In one example, a computer-implemented method comprises: inputting first keywords associated with a first domain corpus; extracting a first keyword of the first keywords; inputting second keywords associated with a second domain corpus; generating an embedded representation of the first keyword via a trained model and generating an embedded representation of the second keywords via the trained model; and scoring a joint embedding affinity associated with a joint embedding.

Abstract: A contextual hashtag generation method, system, and computer program product include receiving content from an online source, identifying a set of contextual indicators for the content, determining an entity-desired outcome for the content, and generating a hashtag for the content using the set of contextual indicators while maximizing the entity-desired outcome.

Abstract: A system and method are shown for transferring weight information to analog non-volatile memory elements wherein the programming pulse duration is directly proportional to the difference in weights. Furthermore, the system and method avoid weight transfers when the weights are already well-matched.

Abstract: Aspects of the invention include a circuit having a power header configured to couple to a power supply and to provide an output voltage. A sense circuit is coupled to the power header to receive the output voltage, the sense circuit including a replica voltage circuit coupled to a replica power header circuit and a transistor, the replica voltage circuit being configured to provide a replicated output voltage in accordance with the output voltage, the replica power header circuit being configured to couple to the power supply and the replicated output voltage to generate a replica current, the transistor being configured to deliver the replica current.

Abstract: A computer-implemented method includes using a cache replacement algorithm to forcefully evict target data from a cache. Using the cache replacement algorithm includes selectively accessing data in the cache. A computer program product includes one or more computer readable storage media and program instructions collectively stored on the one or more computer readable storage media. The program instructions include program instructions to perform the foregoing method. A system includes a processor and logic integrated with the processor, executable by the processor, or integrated with and executable by the processor. The logic is configured to perform the foregoing method.

Abstract: Provided is a reservoir computing system that is miniaturized and has a reduced learning cost. The reservoir computing system uses a reservoir that includes a first optical output section that outputs a first optical signal; a first optical waveguide that propagates the first optical signal output by the first optical output section; an optical receiving section that receives the first optical signal from the first optical waveguide; a storage section that stores received optical data corresponding to the first optical signal and output by the optical receiving section; and a feedback section that applies, to the first optical signal, feedback corresponding to the received optical data stored in the storage section.

Abstract: Information recognition and restructuring includes analyzing electronic media content embedded in an electronic presentment structure presented to a user, and based on the analyzing, detecting portions of the electronic media content previously consumed by the user. The method includes modifying the electronic presentment structure, based on the detecting, to distinguish the electronic media content previously consumed by the user from other portions of the electronic media content.

Abstract: Methods, computer program products, and systems are presented. The method computer program products, and systems can include, for instance: obtaining textual based message data of a messaging system; processing the textual based message data to determine one or more sentiment parameter associated to the textual based message data; and outputting an alert based on a result of the processing.

Abstract: A method includes receiving a plurality of files where each file represents an object. The method further combines selected files of the plurality of files into an aggregation file and stores the aggregation file on low-latency storage media, such as disk storage media. The method represents the aggregation file with a single inode that is stored on the low-latency storage media. The method migrates the aggregate file to high-latency storage media, such as tape storage media. By collecting files into larger aggregation files, the number of files archived is significantly reduced and the files archived are of larger size. The reduced number of files and increased size of the files improves archival and retrieval performance due to reduced metadata operation overhead. A corresponding system and computer program product are also disclosed.

Abstract: Ranking a group of datasets using a computer includes determining a set of target data fields from a set of process documents that indicate user data field preferences. A set of target dataset attributes from a set of data use documents indicate user data scope preferences. A plurality of metadata sets for an associated plurality of datasets the computer determines having a field suitability value exceeding a predetermined suitability threshold value. The FSV represents a degree of similarity between a set of fields associated with said dataset and the set of target data fields. The computer assesses metadata sets with regard to the target attributes and generates a compared attribute score for each candidate dataset. A degree of likelihood is indicated that an associated dataset will have content exhibiting said target dataset attributes. The computer candidate datasets is based on the compared attribute score.