Abstract: A method, system, and computer program product, include receiving a plurality of requests for dynamic context information from a plurality of road segments, determining whether the plurality of road segments are included in a same cluster of road segments in a road network generated by clustering road segments in the road network based on connectivity of the road network; and consolidating the plurality of requests to generate a consolidated request in response to determining that the plurality of road segments are included in the same cluster.

Abstract: A method of processing a wafer is disclosed. The method includes, in some embodiments, causing a relative movement between a cleaning brush and a wafer. During the relative movement, a planar cleaning surface of the cleaning brush is brought into contact with a surface of the wafer to remove contaminants from the surface of the wafer. A first size of the cleaning brush, in a plan view, is larger than a second size of the wafer in the plan view.

Abstract: A MEMS structure includes a substrate, a dielectric layer, a membrane, a backplate, and a blocking layer. The substrate has a through-hole. The dielectric layer is disposed on the substrate and has a cavity in communication with the through-hole. The membrane has at least one vent hole, is embedded in the dielectric layer and together with the dielectric layer defines a first chamber that communicates with the through-hole. The backplate is disposed on the dielectric layer. One end of the blocking layer is embedded in the dielectric layer, and the other end of the blocking layer extends into the cavity; the blocking layer is spatially isolated from the membrane and at least partially overlaps with the at least one vent hole.

Abstract: A semiconductor sensor, comprising a gas-sensing device and an integrated circuit is provided. The gas-sensing device includes a substrate having a sensing area and an interconnection area in the vicinity of the sensing area, an inter-metal dielectric (IMD) layer formed above the substrate in the sensing area and in the interconnection area, and an interconnect structure formed in the interconnection area; further includes a sensing electrode, a second TiO2-patterned portion, and a second Pt-patterned portion on the second TiO2-patterned portion in the sensing area. The interconnect structure includes a tungsten layer buried in the IMD layer, wherein part of a top surface of the tungsten layer is exposed by at least a via. The interconnect structure further includes a platinum layer formed in said at least the via, a TiO2 layer formed on the IMD layer, a first TiO2-patterned portion and a first Pt-patterned portion.

Abstract: A method of generating a predictor to classify data includes: training each of a plurality of first classifiers arranged in a first level on current training data; operating each classifier of the first level on the training data to generate a plurality of predictions; combining the current training data with the predictions to generated new training data; and training each of a plurality of second classifiers arranged in a second level on the new training data. The first classifiers are classifiers of different classifier types, respectively and the second classifiers are classifiers of the different classifier types, respectively.

Abstract: In one embodiment, a computer-implemented method includes receiving training data including a plurality of records, each record having a plurality of attributes. The training data is horizontally parallelized across two or more processing elements. This horizontal parallelizing includes dividing the training data into two or more subsets of records; assigning each subset of records to a corresponding processing element of the two or more processing elements; transmitting each subset of records to its assigned processing element; and sorting, at the two or more processing elements, the two or more subsets of records to two or more candidate leaves of a decision tree. The output from horizontally parallelizing is converted into input for vertically parallelizing the training data. The training data is vertically parallelized across the two or more processing elements. The decision tree is grown based at least in part on the horizontally parallelizing, the converting, and the vertically parallelizing.

Abstract: In one embodiment, a computer-implemented method includes receiving training data including a plurality of records, each record having a plurality of attributes. The training data is horizontally parallelized across two or more processing elements. This horizontal parallelizing includes dividing the training data into two or more subsets of records; assigning each subset of records to a corresponding processing element of the two or more processing elements; transmitting each subset of records to its assigned processing element; and sorting, at the two or more processing elements, the two or more subsets of records to two or more candidate leaves of a decision tree. The output from horizontally parallelizing is converted into input for vertically parallelizing the training data. The training data is vertically parallelized across the two or more processing elements. The decision tree is grown based at least in part on the horizontally parallelizing, the converting, and the vertically parallelizing.

Abstract: A semiconductor sensor, comprising a gas-sensing device and an integrated circuit electrically connected to the gas-sensing device, is provided. The gas-sensing device includes a substrate having a sensing area and an interconnection area in the vicinity of the sensing area, an inter-metal dielectric (IMD) layer formed above the substrate in the sensing area and in the interconnection area, and an interconnect structure formed in the interconnection area. The interconnect structure includes a tungsten layer buried in the IMD layer, wherein part of a top surface of the tungsten layer is exposed by at least a via. The interconnect structure further includes a platinum layer formed in said at least the via, wherein the platinum (Pt) layer directly contacts the top surface of the tungsten layer.

Abstract: A method for non-invasive glucose monitoring includes the following steps. At least one ray of light is emitted from at least one light source. The light emitted from the light source is leaded into an eyeball and focused on the eyeball through a first beam splitter. The reflected light reflected from the eyeball is transmitted through the first beam splitter to a set of photo detectors. Optical angular information and energy information of the reflected light transmitted to the set of photo detectors are measured. Optical angular difference and energy difference resulting from the light emitted from the light source and the reflected light transmitted to the set of photo detectors are obtained. Glucose information is obtained by analyzing the optical angular difference and the energy difference. Since glucose information has a corresponding relationship with blood glucose information, blood glucose information may be obtained.

Abstract: A system, a computer readable storage medium, and a method for detecting insurance fraud or for comparing references use visual analytics-based techniques. The method can include identifying a scratch and a scratch location on a vehicle in a 3-dimensional rendering in comparison to a base image or in comparison to a stored image in a database, comparing one or more features of the scratch such as color features of the scratch in the scratch location in comparison to the scratch location or comparing one or more texture features of the scratch in the scratch location in comparison to the scratch location in the base image or the stored image. The method can further generate a similarity calculation based on the one or more comparisons for the scratch, the scratch location, the one or more scratch features of the scratch and presents a result in response to the similarity calculation.

Abstract: A method includes providing a structure having a substrate and a fin protruding from the substrate; forming a dummy gate stack over the fin; forming a gate spacer on sidewalls of the dummy gate stack; removing the dummy gate stack using a radical etch process, resulting in a gate trench; and forming a metal gate stack in the gate trench.

Abstract: A fatigue alarm apparatus having a fatigue detector, a control unit and a stimulus alarm generator is illustrated. The fatigue detector detects whether a driver is drowsy. The control unit is electrically connected to the fatigue detector and the stimulus alarm generator. When the fatigue detector detects that the driver is drowsy, the stimulus alarm generator controlled by the control unit to generate two stimulus alarms having a rest time period therebetween, and then, after an interval period elapses, the control unit determines whether an alarming procedure of generating the two stimulus alarms is to be interrupted or repeated, wherein the rest time period is less than the interval period.

Abstract: A method for fabricating an integrated circuit is provided. The method includes depositing a first polish stop layer above a memory device, in which the first polish stop layer has a first portion over the memory device and a second portion that is not over the memory device; removing the second portion of the first polish stop layer; depositing an inter-layer dielectric layer over the first polish stop layer after removing the second portion of the first polish stop layer; and polishing the inter-layer dielectric layer until reaching the first portion of the first polish stop layer.

Abstract: A mechanism is provided in a data processing system for distributed tree learning. A source processing instance distributes data record instances to a plurality of model update processing items. The plurality of model update processing items determine candidate leaf splitting actions in a decision tree in parallel based on the data record instances. The plurality of model update processing items send the candidate leaf splitting actions to a plurality of conflict resolve processing items. The plurality of conflict resolve processing items identifies conflict leaf splitting actions. The plurality of conflict resolve processing items applies tree structure changes to the decision tree in the plurality of model update processing items.

Abstract: A mechanism is provided in a data processing system for distributed tree learning. A source processing instance distributes data record instances to a plurality of model update processing items. The plurality of model update processing items determine candidate leaf splitting actions in a decision tree in parallel based on the data record instances. The plurality of model update processing items send the candidate leaf splitting actions to a plurality of conflict resolve processing items. The plurality of conflict resolve processing items identifies conflict leaf splitting actions. The plurality of conflict resolve processing items applies tree structure changes to the decision tree in the plurality of model update processing items.

Abstract: A method includes forming a first dielectric layer over a wafer, etching the first dielectric layer to form an opening, filling a tungsten-containing material into the opening, and performing a Chemical Mechanical Polish (CMP) on the wafer. After the CMP, a cleaning is performed on the wafer using a weak base solution.

Abstract: The present disclosure, in some embodiments, relates to a brush cleaning apparatus. The brush cleaning apparatus includes a wafer support configured to support a wafer. The brush cleaning apparatus also includes a cleaning brush including a porous material coupled to a core material. An uppermost surface of the porous material defines a planar cleaning surface. A first nozzle is configured to apply a first cleaning liquid directly between the wafer and the planar cleaning surface of the cleaning brush.

Abstract: A method includes forming a first dielectric layer over a wafer, etching the first dielectric layer to form an opening, filling a tungsten-containing material into the opening, and performing a Chemical Mechanical Polish (CMP) on the wafer. After the CMP, a cleaning is performed on the wafer using a weak base solution.

Abstract: A brush cleaning apparatus includes a wafer support configured to support a wafer, and at least one cleaning brush moveable relative to the wafer support. The at least one cleaning brush has opposite first and second sides, and, on the first side, a planar cleaning surface configured to come into contact with the wafer supported by the wafer support to remove contaminants from the wafer.

Abstract: Some embodiments relate to a carrier head. The carrier head includes a housing configured to enclose a wafer, wherein the housing includes a retaining ring recess configured to circumferentially surround the wafer. A retaining ring, which includes a first ring-shaped layer and a second ring-shaped layer, is disposed in the retaining ring recess. The second ring-shaped layer is disposed deeper in the retaining ring recess than the first ring-shaped layer and separates the first ring-shaped layer from a bottom of the retaining ring recess. A hardness of the second ring-shaped layer is less than a hardness of the first ring-shaped layer.