Abstract: A method for operating a data processing system to generate an estimate of radiative contamination at nodes in an RF circuit characterized by a plurality of circuit elements connected by metal traces on a circuit board are disclosed. The data processing system to receive information specifying a coupled radiation matrix based on the metal traces and a simulation of an RF circuit with the components connected by non-radiating nodes. The data processing system generates a coupled power list for at least one node of the model, each entry in the coupled power list includes a coupled power value indicating a power level received by EM radiation from another of the nodes.

Abstract: Techniques are provided herein for contextual embedding of features of operational logs or network traffic for anomaly detection based on sequence prediction. In an embodiment, a computer has a predictive recurrent neural network (RNN) that detects an anomalous network flow. In an embodiment, an RNN contextually transcodes sparse feature vectors that represent log messages into dense feature vectors that may be predictive or used to generate predictive vectors. In an embodiment, graph embedding improves feature embedding of log traces. In an embodiment, a computer detects and feature-encodes independent traces from related log messages. These techniques may detect malicious activity by anomaly analysis of context-aware feature embeddings of network packet flows, log messages, and/or log traces.

Abstract: Elemental concentrations in subterranean formations may be determined using neutron spectroscopy. For example, neutrons may be emitted by a downhole tool into the formation and produce gamma rays via inelastic scattering of fast neutrons or capture of slow neutrons. The borehole surrounding a downhole tool may introduce artifacts in the neutron spectroscopy measurement. Embodiments of the present disclosure are directed to techniques that reduce artifacts signals in downhole tools that include one or multiple detectors based at least in part on the inelastic and capture measurements.

Abstract: A hot-standby redundancy control system is disclosed. In an embodiment, the system includes multiple primary control apparatuses, each respectively running a process of the hot-standby redundancy control system; a backup control apparatus pool including at least one industrial personal computer (IPC), with multiple virtual control apparatuses being established on at least one IPC, and the multiple virtual control apparatuses established being in one-to-one correspondence with the multiple primary control apparatuses, and each respective multiple virtual control apparatus running a same process as a respective corresponding primary control apparatus; a control bus, to connect one or more of the t least one IPC in the multiple primary control apparatuses and the backup control apparatus pool; and a field bus, to connect one or more of the at least one IPC in the multiple primary control apparatuses and the backup control apparatus pool and multiple field apparatuses.

Abstract: A method for generating a digital identity is provided, including: extracting a first preset number of short tandem repeat STRs and relevant information of each STR from whole genome data; generating a single STR digital code corresponding to each STR according to the relevant information of each STR, to obtain a plurality of single STR digital codes; performing sequence transformation on each single STR digital code with a preset rule, and generating a target STR digital code according to the single STR digital code after the sequence transformation; generating summary information of the target STR digital code, and determining the summary information as summary information of the STR to which the target STR digital code belongs; and determining the summary information of the STR as the generated digital identity.

Abstract: Techniques are described for closing one or more connections with a database management system (DBMS) as part of a planned operation such as maintenance and load balancing. In an embodiment, a request is received for a planned operation of a database server of a DBMS that requires releasing and migrating database sessions. Over a database connection with the database server, the DBMS or a Client-Driver on the client system that interfaces with the DBMS, may receive request(s) from the client system. Based on the request(s), a safe point, for the database connection with the database server, at which the database connection of the client system with the database server can be closed without affecting the success of execution of next one or more command that the client system intended to request the DBMS to execute on the database server, is determined.

Abstract: A wideband vibration suppression device utilizing properties of a sonic black hole, comprising: a vibration absorber (101) comprising a uniform portion (1011) of a fixed thickness and a conical portion (1012) integrally connected to the uniform portion (1011), the conical portion (1012) extending from the junction in such a manner that the thickness thereof gradually decreases from the thickness (d1) of the uniform portion (1011) to a predetermined thickness (d2); and a damping layer (102) attached to the conical portion (1012) of the vibration absorber (101).

Abstract: Provided are methods for diagnosing and predicting the outcome of a breast cancer patient or related cancers. The methods include determining expression levels of a plurality of biomarkers (selected genes) related to immune function in a biological sample, such as tumor tissues or a body fluid such as blood, from the patient. The expression levels of the biomarkers are used to derive an index which can be used as an indicator predictive of cancer patient outcome. Overexpression of a plurality of biomarkers of the invention can be used to generate a score or value which has been demonstrated herein to be indicative of a good or better patient outcome. The index, generated by the methods of the invention can also be used to stratify cancer subtypes, and also can be combined with conventional clinical parameters to better inform clinical decisions.

Abstract: A method for determining a fractional volume of at least one component of a formation includes entering into a computer a number of detected radiation events resulting from imparting neutrons into the formation at an energy level of at least 1 million electron volts (MeV). The detected radiation events correspond to at least one of an energy level of the imparted neutrons and thermal or epithermal energy neutrons. A measurement of at least one additional petrophysical parameter of the formation is made. The at least one additional petrophysical parameter measurement and at least one of a fast neutron cross-section and a thermal neutron cross-section determined from the detected radiation events are used in the computer to determine the fractional volume of the at least one component of the formation. In another embodiment, the fast neutron cross-section and the thermal neutron cross-section may be used on combination to determine the fractional volume.

Abstract: In accordance with an embodiment, the system enables access to a sharded database using a cache and a shard topology. A shard-aware client application connecting to a sharded database can use a connection pool (e.g., a Universal Connection Pool, UCP), to store or access connections to different shards or chunks of the sharded database within a shared pool. As new connections are created, a shard topology layer can be built at the database driver layer, which learns and caches shard key ranges to locations of shards. The shard topology layer enables subsequent connection requests from a client application to use a fast key path access to the appropriate shard or chunk.

Abstract: Techniques are described herein for predicting disk drive failure using a machine learning model. The framework involves receiving disk drive sensor attributes as training data, preprocessing the training data to select a set of enhanced feature sequences, and using the enhanced feature sequences to train a machine learning model to predict disk drive failures from disk drive sensor monitoring data. Prior to the training phase, the RNN LSTM model is tuned using a set of predefined hyper-parameters. The preprocessing, which is performed during the training and evaluation phase as well as later during the prediction phase, involves using predefined values for a set of parameters to generate the set of enhanced sequences from raw sensor reading. The enhanced feature sequences are generated to maintain a desired healthy/failed disk ratio, and only use samples leading up to a last-valid-time sample in order to honor a pre-specified heads-up-period alert requirement.

Abstract: An instance node management method is disclosed, which includes: creating a primary instance node for an application hosted on a cloud platform to run an instance of the application; determining, according to running start information of the primary instance node, a minimum resource configuration specification required by a secondary instance node of the application, where the secondary instance node is configured to replace the primary instance node when the primary instance node is faulty; and creating the secondary instance node according to the minimum resource configuration specification. The instance node management method as provided can reduce occupancy of a spare resource and increase an application hosting capability and scale on the cloud platform.

Abstract: A method for determining a fractional volume of at least one component of a formation includes entering into a computer a number of detected radiation events resulting from imparting neutrons into the formation at an energy level of at least 1 million electron volts (MeV). The detected radiation events correspond to at least one of an energy level of the imparted neutrons and thermal or epithermal energy neutrons. A measurement of at least one additional petrophysical parameter of the formation is made. The at least one additional petrophysical parameter measurement and at least one of a fast neutron cross-section and a thermal neutron cross-section determined from the detected radiation events are used in the computer to determine the fractional volume of the at least one component of the formation. In another embodiment, the fast neutron cross-section and the thermal neutron cross-section may be used on combination to determine the fractional volume.

Abstract: Techniques are provided herein for contextual embedding of features of operational logs or network traffic for anomaly detection based on sequence prediction. In an embodiment, a computer has a predictive recurrent neural network (RNN) that detects an anomalous network flow. In an embodiment, an RNN contextually transcodes sparse feature vectors that represent log messages into dense feature vectors that may be predictive or used to generate predictive vectors. In an embodiment, graph embedding improves feature embedding of log traces. In an embodiment, a computer detects and feature-encodes independent traces from related log messages. These techniques may detect malicious activity by anomaly analysis of context-aware feature embeddings of network packet flows, log messages, and/or log traces.

Abstract: Techniques are provided herein for contextual embedding of features of operational logs or network traffic for anomaly detection based on sequence prediction. In an embodiment, a computer has a predictive recurrent neural network (RNN) that detects an anomalous network flow. In an embodiment, an RNN contextually transcodes sparse feature vectors that represent log messages into dense feature vectors that may be predictive or used to generate predictive vectors. In an embodiment, graph embedding improves feature embedding of log traces. In an embodiment, a computer detects and feature-encodes independent traces from related log messages. These techniques may detect malicious activity by anomaly analysis of context-aware feature embeddings of network packet flows, log messages, and/or log traces.

Abstract: Techniques are provided herein for contextual embedding of features of operational logs or network traffic for anomaly detection based on sequence prediction. In an embodiment, a computer has a predictive recurrent neural network (RNN) that detects an anomalous network flow. In an embodiment, an RNN contextually transcodes sparse feature vectors that represent log messages into dense feature vectors that may be predictive or used to generate predictive vectors. In an embodiment, graph embedding improves feature embedding of log traces. In an embodiment, a computer detects and feature-encodes independent traces from related log messages. These techniques may detect malicious activity by anomaly analysis of context-aware feature embeddings of network packet flows, log messages, and/or log traces.

Abstract: In accordance with an embodiment, the system enables access to a sharded database using a cache and a shard topology. A shard-aware client application connecting to a sharded database can use a connection pool (e.g., a Universal Connection Pool, UCP), to store or access connections to different shards or chunks of the sharded database within a shared pool. As new connections are created, a shard topology layer can be built at the database driver layer, which learns and caches shard key ranges to locations of shards. The shard topology layer enables subsequent connection requests from a client application to use a fast key path access to the appropriate shard or chunk.

Abstract: The disclosed embodiments relate to a system for validating a prognostic-surveillance mechanism, which detects anomalies that arise during operation of a computer system. During operation, the system obtains telemetry data comprising a set of raw signals gathered from sensors in the computer system during operation of the computer system, wherein the telemetry signals are gathered over a monitored time period. Next, for each raw signal in the set of raw signals, the system decomposes the raw signal into deterministic and stochastic components. The system then generates a corresponding set of synthesized signals based on the deterministic and stochastic components of the raw signals, wherein the synthesized signals are generated for a simulated time period, which is longer than the monitored time period. Finally, the system uses the set of synthesized signals to validate one or more performance metrics of the prognostic-surveillance mechanism.

Abstract: Elemental concentrations in subterranean formations may be determined using neutron spectroscopy. For example, neutrons may be emitted by a downhole tool into the formation and produce gamma rays via inelastic scattering of fast neutrons or capture of slow neutrons. The borehole surrounding a downhole tool may introduce artifacts in the neutron spectroscopy measurement. Embodiments of the present disclosure are directed to techniques that reduce artifacts signals in downhole tools that include one or multiple detectors based at least in part on the inelastic and capture measurements.

Abstract: A method for evaluating a formation includes determining a number of detected gamma rays resulting from imparting neutrons into a formation. The detected gamma rays are each characterized by an energy level thereof. The gamma rays are detected at a first distance from a position of imparting the neutrons into the formation. Those of the detected gamma rays attributable to neutron capture by hydrogen nuclei are removed from the number of detected gamma rays. The number of detected gamma rays having hydrogen neutron capture gamma rays removed therefrom are used to calculate a property of the formation.