Abstract: A method includes receiving a first data set comprising embeddings of first and second types, generating a fixed adjacency matrix from the first dataset, and applying a first stochastic binary mask to the fixed adjacency matrix to obtain a first subgraph of the fixed adjacency matrix. The method also includes processing the first subgraph through a first layer of a graph convolutional network (GCN) to obtain a first embedding matrix, and applying a second stochastic binary mask to the fixed adjacency matrix to obtain a second subgraph of the fixed adjacency matrix. The method includes processing the first embedding matrix and the second subgraph through a second layer of the GCN to obtain a second embedding matrix, and then determining a plurality of gradients of a loss function, and modifying the first stochastic binary mask and the second stochastic binary mask using at least one of the plurality of gradients.

Abstract: An encoding system may be provided. The encoding system may comprise a first stage and a second stage. The first stage may be configured to receive a first input, decode the first input, and produce a first output comprising the decoded first input. The second stage may be configured to receive a second input, receive the first output from the first stage, and convert the first input and the second input from a first coding system to a second coding system based on the second input and the first output. The second stage may produce a second output comprising the converted first input and the converted second input.

Abstract: A system includes a mask. The system further includes a pellicle frame attached to the mask. The pellicle frame includes a check valve, wherein the check valve is configured to permit gas flow from a first side of the pellicle from to a second side of the pellicle frame. The pellicle frame further includes a flat bottom surface having only a single recess therein, wherein the flat bottom surface is free of an adhesive. The system further includes a gasket within the single recess.

Abstract: Systems, methods, and computer program products for multi-domain ensemble learning based on multivariate time sequence data are provided. A method may include receiving multivariate sequence data. At least a portion of the multivariate sequence data may be inputted into a plurality of anomaly detection models to generate a plurality of scores. The multivariate sequence data may be combined with the plurality of scores to generate combined intermediate data. The combined intermediate data may be inputted into a combined ensemble model to generate an output score. In response to determining that the output score satisfies a threshold, at least one of an alert may be communicated to a user device, the multivariate sequence data may be inputted into the feature-domain ensemble model to generate a feature importance vector, or at least one of a model-domain, a time-domain, a feature-domain, or the combined ensemble model may be updated.

Abstract: A voltage detection device is provided. The voltage detection device includes a first voltage divider circuit, a comparison circuit, and a second voltage divider circuit. The first voltage divider circuit is configured to receive an input voltage and output a comparison voltage according to the input voltage. The comparison circuit is configured to receive the comparison voltage to compare the comparison voltage with a reference voltage and determine whether to change a trigger signal according to a comparison result. The second voltage divider circuit is configured to receive the input voltage. When the input voltage is greater than or equal to a predetermined voltage value, the second voltage divider circuit and the first voltage dividing circuit form a parallel structure to pull down the comparison voltage.

Abstract: A method includes receiving a first data set comprising embeddings of first and second types, generating a fixed adjacency matrix from the first dataset, and applying a first stochastic binary mask to the fixed adjacency matrix to obtain a first subgraph of the fixed adjacency matrix. The method also includes processing the first subgraph through a first layer of a graph convolutional network (GCN) to obtain a first embedding matrix, and applying a second stochastic binary mask to the fixed adjacency matrix to obtain a second subgraph of the fixed adjacency matrix. The method includes processing the first embedding matrix and the second subgraph through a second layer of the GCN to obtain a second embedding matrix, and then determining a plurality of gradients of a loss function, and modifying the first stochastic binary mask and the second stochastic binary mask using at least one of the plurality of gradients.

Abstract: A power supply phase doubling system includes a pulse width modulation (PWM) controller and first and second phase doubling chips. The PWM controller outputs a PWM signal. The first phase doubling chip is operated at a power supply voltage and has a first PWM output pin to generate a first control signal and a second control signal according to the PWM signal, and generates a first output signal according to the first control signal. The second phase doubling chip is operated at the power supply voltage, has a second PWM output pin, and is configured to generate a second output signal according to the second control signal. The first and second phase doubling chips are respectively switched between a master mode and a slave mode according to a voltage level of the first PWM output pin and a voltage level of the second PWM output pin.

Abstract: Disclosed are a system, method, and computer program product for user network activity anomaly detection. The method includes generating a multilayer graph from network resource data, and generating an adjacency matrix associated with each layer of the multilayer graph to produce a plurality of adjacency matrices. The method further includes assigning a weight to each adjacency matrix to produce a plurality of weights, and generating a merged single layer graph by merging the plurality of layers based on a weighted sum of the plurality of adjacency matrices using the plurality of weights. The method further includes generating a set of anomaly scores by generating, for each node in the merged single layer graph, an anomaly score. The method further includes determining a set of anomalous users based on the set of anomaly scores, detecting fraudulent network activity based on the set of anomalous users, and executing a fraud mitigation process.

Abstract: Described are a system, method, and computer program product for user network activity anomaly detection. The method includes receiving network resource data associated with network resource activity of a plurality of users and generating a plurality of layers of a multilayer graph from the network resource data. Each layer of the plurality of layers may include a plurality of nodes, which are associated with users, connected by a plurality of edges, which are representative of node interdependency. The method also includes generating a plurality of adjacency matrices from the plurality of layers and generating a merged single layer graph based on a weighted sum of the plurality of adjacency matrices. The method further includes generating anomaly scores for each node in the merged single layer graph and determining a set of anomalous users based on the anomaly scores.

Abstract: Described are a system, method, and computer program product for user network activity anomaly detection. The method includes receiving network resource data associated with network resource activity of a plurality of users and generating a plurality of layers of a multilayer graph from the network resource data. Each layer of the plurality of layers may include a plurality of nodes, which are associated with users, connected by a plurality of edges, which are representative of node interdependency. The method also includes generating a plurality of adjacency matrices from the plurality of layers and generating a merged single layer graph based on a weighted sum of the plurality of adjacency matrices. The method further includes generating anomaly scores for each node in the merged single layer graph and determining a set of anomalous users based on the anomaly scores.

Abstract: Provided is a system for event forecasting using a graph-based machine-learning model that includes at least one processor programmed or configured to receive a dataset of data instances, where each data instance comprises a time series of data points, detect a plurality of motifs representing a plurality of events in the dataset of data instances using a matrix profile-based motif detection technique, generate a bipartite graph representation of the plurality of motifs in a time sequence, and generate a machine-learning model based on the bipartite graph representation of the plurality of motifs in the time sequence, where the machine-learning model is configured to provide an output and the output includes a prediction of whether an event will occur during a specified time interval. Methods and computer program products are also provided.

Abstract: Methods for detecting insider threats are disclosed. A method includes collecting server access data and application access data, based on the server access data and the application access data, determining nearest neighbors of an employee, and based on the nearest neighbors of the employee, determining a peer group of the employee, determining an average rank distance (ARD) of the nearest neighbors based on a ranking of the nearest neighbors in a plurality of time periods, identifying ARD gaps between the nearest neighbors, and generating scores corresponding to the ARD gaps between the nearest neighbors. One or more employees are identified that represent an internal threat to an organization based on the scores corresponding to the ARD gaps.

Abstract: Systems, methods, and computer program products for predicting user preference of items in an image process image data associated with a single image with a first branch of a neural network to produce an image embedding, the single image including a set of multiple items; process a user identifier of a user with a second branch of the neural network to produce a user embedding; concatenate the image embedding with the user embedding to produce a concatenated embedding; process the concatenated embedding with the neural network to produce a joint embedding; and generate a user preference score for the set of multiple items from the neural network based on the joint embedding, the user preference score including a prediction of whether the user prefers the set of multiple items.

Abstract: Methods for detecting insider threats are disclosed. A method includes collecting server access data and application access data, based on the server access data and the application access data, determining nearest neighbors of an employee, and based on the nearest neighbors of the employee, determining a peer group of the employee, determining an average rank distance (ARD) of the nearest neighbors based on a ranking of the nearest neighbors in a plurality of time periods, identifying ARD gaps between the nearest neighbors, and generating scores corresponding to the ARD gaps between the nearest neighbors. One or more employees are identified that represent an internal threat to an organization based on the scores corresponding to the ARD gaps.

Abstract: A voltage converter is provided. The voltage converter includes a charge pump and a switching circuit. The charge pump includes an input capacitor, where two ends of the input capacitor are electrically connected to an input end and a ground end respectively. The switching circuit includes a first switch, a second switch, a third switch, and a fourth switch, and these switches are connected in series. The first switch is electrically connected to the input end, the fourth switch is electrically connected to the ground end, a capacitor is connected in parallel with the second switch and the third switch, and an output end is electrically connected between the second switch and the third switch.

Abstract: The present application provides a wireless charging module and a modularized sensing device therewith. A first charging unit is disposed around the inside of the wireless charging module for wirelessly charging a sensing module. Without introducing a transmission line as an electricity transmitting medium, the sensing module can be directly disposed inside and charged by the wireless charging module, which improves flexibility of implementation. Further, since the first charging unit is configured to surround the inside of the wireless charging module shaped as a bowl structure, the sensing module can be simply disposed into the bowl structure and wirelessly charged by the surrounding first charging units, which saves the inconvenience of setting the sensing module to a specific location or area for stable wireless charging and increases effective charging angles.

Abstract: The present application provides a wireless charging module and a modularized sensing device therewith. A first charging unit is disposed around the inside of the wireless charging module for wirelessly charging a sensing module. Without introducing a transmission line as an electricity transmitting medium, the sensing module can be directly disposed inside and charged by the wireless charging module, which improves flexibility of implementation. Further, since the first charging unit is configured to surround the inside of the wireless charging module shaped as a bowl structure, the sensing module can be simply disposed into the bowl structure and wirelessly charged by the surrounding first charging units, which saves the inconvenience of setting the sensing module to a specific location or area for stable wireless charging and increases effective charging angles.

Abstract: A clip cover for a C-type light string is an auxiliary accessory exclusively used for a C-type light string, and mainly includes a wire cover constituted by first front, rear covers, a lamp cover constituted by second front, rear covers and a buddle cover. Electric wire segments at the two sides of one C-type lamp of the light string are clipped inside the wire cover by engaging the first front cover with the first rear cover. Thereafter, the electric wire is pulled downward from the wire cover a certain length, allowing the C-type lamp to be hung perpendicularly below the wire cover; the lamp socket of the C-type lamp together with a small section of bent electric wire is further clipped inside the lamp cover through the second front, rear covers of the lamp cover; and the bundle cover is finally put around the lamp cover tightly.

Abstract: Disclosed is a bubble lamp for quasi bubble flow. A buffer made of a transparent material having an excellent heat conduction effect and impact moderating performance is configured on the top of the bulb configured on the main body of the bubble lamp or on the bottom of the glass tube configured on the upper cover, and the casing made of a transparent and no-easily-broken material is put around the glass tube, thereby abutting the glass tube against the top of the bulb across the buffer when the upper cover is covered on and coupled to the main body, allowing both the glass-made bulb and glass tube to be buffered by the butter and thus to block or decrease an impact when a contact force is too large upon assembly, and preventing the bulb or glass tube from being broken.

Abstract: A clip cover for a C-type light string is an auxiliary accessory exclusively used for a C-type light string, and mainly includes a wire cover constituted by first front, rear covers, a lamp cover constituted by second front, rear covers and a buddle cover. Electric wire segments at the two sides of one C-type lamp of the light string are clipped inside the wire cover by engaging the first front cover with the first rear cover. Thereafter, the electric wire is pulled downward from the wire cover a certain length, allowing the C-type lamp to be hung perpendicularly below the wire cover; the lamp socket of the C-type lamp together with a small section of bent electric wire is further clipped inside the lamp cover through the second front, rear covers of the lamp cover; and the bundle cover is finally put around the lamp cover tightly.