Abstract: A method for receiving training data for training a neural network to perform a machine learning task and for searching for, using the training data, an optimized neural network architecture for performing the machine learning task is described. Searching for the optimized neural network architecture includes: maintaining population data; maintaining threshold data; and repeatedly performing the following operations: selecting one or more candidate architectures from the population data; generating a new architecture from the one or more selected candidate architectures; for the new architecture: training a neural network having the new architecture until termination criteria for the training are satisfied; and determining a final measure of fitness of the neural network having the new architecture after the training; and adding data defining the new architecture and the final measure of fitness for the neural network having the new architecture to the population data.

Abstract: Methods and systems for learning policies using sparse and underspecified rewards. One of the methods includes training the policy jointly with an auxiliary reward function having a plurality of auxiliary reward parameters, the auxiliary reward function being configured to map, in accordance with the auxiliary reward parameters, trajectory features of at least a trajectory to an auxiliary reward value that indicates how well the trajectory performed a task in response to a context input.

Abstract: A semiconductor device includes a semiconductor substrate, a gate dielectric, a gate electrode, and a pair of source/drain regions. The gate dielectric is disposed in the semiconductor substrate having an upper boundary lower than an upper surface of the semiconductor substrate, and an upper surface flush with the upper surface of the semiconductor substrate. The gate electrode is disposed over the gate dielectric having a first section over the upper boundary of the gate dielectric and a second section over the upper surface of the gate dielectric. The second section partially covers and partially exposes the upper surface of the gate dielectric. The pair of source/drain regions are disposed on opposing sides of the gate dielectric.

Abstract: Techniques for divisive clustering of a dataset to identify consumer choice patterns are described herein. The techniques include accessing a data source having a dataset to be analyzed and obtaining a feature list upon which the dataset is clustered. The dataset is hierarchically clustered using divisive clustering by estimating a conditional probability of stickiness for each feature of the feature list within the dataset. The feature having the greatest probability of stickiness is selected and used to split the dataset into clusters based on the feature. Then each cluster or branch of the dataset is recursively clustered using the same technique of estimating the probability of stickiness for each of the remaining features, selecting the feature with the highest probability of stickiness, and dividing the remaining dataset into clusters based on that feature. A nested logit model is generated using the hierarchical clustering and used to identify consumer choice patterns.

Abstract: Various embodiments of the present disclosure are directed towards a semiconductor device. The semiconductor device comprises a source region and a drain region in a substrate and laterally spaced. A gate stack is over the substrate and between the source region and the drain region. The drain region includes two or more first doped regions having a first doping type in the substrate. The drain region further includes one or more second doped regions in the substrate. The first doped regions have a greater concentration of first doping type dopants than the second doped regions, and each of the second doped regions is disposed laterally between two neighboring first doped regions.

Abstract: The present invention discloses a special hydraulic press for bulging pressing forming of an automobile axle housing and a pressing forming method. A hydraulic press body is a closed frame consisting of an upper cross beam, a lower cross beam, four stand columns I, II, III, and IV, four vertical pull rods I, II, III, and IV, and four transverse pull rods I, II, III, and IV, and is used for balancing a deformation force up to 3000 T in the vertical direction or the horizontal direction in the pressing forming process. During pressing forming, the downward speeds of a main hydraulic cylinder, small left and right vertical hydraulic cylinders are precisely controlled to prevent the left end, the right end and the middle of the tube blank from bending.

Abstract: A wafer support device includes a susceptor, at least one lift pin, at least one lift pin support base and at least one pad. The susceptor has a bottom surface and a top surface configured to support a wafer. The susceptor has at least one through hole extending between the bottom surface and the top surface. The lift pin is at least partially telescopically received in the through hole of the susceptor. The lift pin support base has at least one coupling feature thereon. The pad is detachably coupled with the coupling feature and supports the lift pin.

Abstract: A method for receiving training data for training a neural network to perform a machine learning task and for searching for, using the training data, an optimized neural network architecture for performing the machine learning task is described. Searching for the optimized neural network architecture includes: maintaining population data; maintaining threshold data; and repeatedly performing the following operations: selecting one or more candidate architectures from the population data; generating a new architecture from the one or more selected candidate architectures; for the new architecture: training a neural network having the new architecture until termination criteria for the training are satisfied; and determining a final measure of fitness of the neural network having the new architecture after the training; and adding data defining the new architecture and the final measure of fitness for the neural network having the new architecture to the population data.

Abstract: The present invention provides a method for improving the construction process of a wide-base radial truck tire, and the improvement of the tire construction process is realized by the design of carcass ply, the tire belt drum and the tire semi-finished components. The invention solves the problem of carcass bending, and provides theoretical guidance and technical support for improving the performance of wide-angle radial tires. The design method of the belt drum of the tire molding machine effectively controls the cords elongation of the belt. In addition, the problem of uneven force in the cords of the belt is eliminated, which dispels the abnormal flow of the wide-base tire shoulder rubber to the center of the crown during the shaping process, thus improving the durability and operability of the tire.

Abstract: A semiconductor device includes a semiconductor substrate, a gate dielectric, a gate electrode, a pair of source/drain regions, a pair of first well regions, a second well region, a pair of contact regions and a pair of third well regions. The gate dielectric is disposed in the semiconductor substrate having a concave profile that defines an upper boundary lower than an upper surface of the semiconductor substrate. The gate electrode is disposed over the gate dielectric. The pair of source/drain regions are disposed on opposing sides of the gate dielectric. The pair of first well regions are disposed under the pair of source/drain regions. The second well region is disposed between the pair of first well regions. The pair of contact regions are disposed on opposing sides of the pair of source/drain regions. The pair of third well regions are disposed under the pair of contact regions.

Abstract: The present disclosure provides a method for preparing nitrogen oxide gas sensor based on sulfur-doped graphene. The method includes: 1) providing graphene and a micro heater platform substrate, and transferring the graphene onto the micro heater platform substrate; 2) putting the micro heater platform substrate covered with the graphene into a chemical vapor deposition reaction furnace; 3) performing gas feeding and exhausting treatment to the reaction furnace by using inert gas; 4) simultaneously feeding inert gas and hydrogen gas into the reaction furnace at a first temperature; 5) feeding inert gas, hydrogen gas and sulfur source gas into the reaction furnace at a second temperature for reaction to perform sulfur doping to the graphene (21); and 6) stopping feeding the sulfur source gas, and performing cooling in a hydrogen gas and insert gas shielding atmosphere.

Abstract: A non-volatile memory structure including a substrate, a stacked structure, a conductive pillar, a channel layer, a charge storage structure, and a second dielectric layer is provided. The stacked structure is disposed on the substrate and has an opening. The stacked structure includes first conductive layers and first dielectric layers alternately stacked. The conductive pillar is disposed in the opening. The channel layer is disposed between the stacked structure and the conductive pillar. The charge storage structure is disposed between the stacked structure and the channel layer. The second dielectric layer is disposed between the channel layer and the conductive pillar. The non-volatile memory structure can effectively improve the electrical performance and the reliability of the memory device.

Abstract: A processor device has a CPU cooperating with an input device and an output device, under control of stored instructions, and is arranged to receive service requests at the input device, assign service requests received in successive time periods to respective batches of requests; access stored service provider data to identify available service providers from among a pool of service providers; after completing the assignment of service requests to a batch, perform a matching process to endeavour to match each service request of the batch of requests to a service provider; and for each service provider to whom a match is made, output a notification of the respective potential match from the output device.

Abstract: A method, system and computer program product for automatically distributing tasks within a group includes identifying, by one or more processors, first data associated with each member of a group. The one or more processors identify second data associated with demands for the group, calculate a total work time for each member of the group using the first data, identify conflicts between the total work time for each member of the group and the second data, and based on the identified conflict, distribute tasks among members of the group.

Abstract: A secured virtual container is enabled to securely store personal data corresponding to a user, where such data is inaccessible to processes running outside the secured virtual container. The secured virtual container may also include an execution environment for a machine learning model where the model is securely stored and inaccessible. Personal data may be feature engineered and provided to the machine learning model for training purposes and/or to generate inference values corresponding to the user data. Inference values may thereafter be relayed by a broker application from the secured virtual container to applications external to the container. Applications may perform hyper-personalization operations based at least in part on received inference values. The broker application may enable external applications to subscribe to notifications regarding availability of inference values. The broker may also provide inference values in response to a query.

Abstract: A semiconductor device is provided. The semiconductor device comprises a substrate, a gate, a first doped region and a second doped region. The gate is over the substrate. The first doped region and the second doped region are in the substrate. The first doped region and the second doped region are of a same conductivity type and separated by the gate. The length of the first doped region is greater than a length of the second doped region in a direction substantially perpendicular to a channel length defined between the first doped region and the second doped region.

Abstract: A method may add whey to high calcium papermaking industrial wastewater blended and pre-acidification treatment to promote anaerobic reaction and inhibit calcification. The method includes, before anaerobic treatment of papermaking wastewater, mixing whey wastewater with the papermaking wastewater. The method also includes pre-treating the papermaking wastewater by acidification to stabilize the papermaking wastewater prior to entering an anaerobic reaction system.

Abstract: A method may add whey to high calcium papermaking industrial wastewater blended and pre-acidification treatment to promote anaerobic reaction and inhibit calcification. The method includes, before anaerobic treatment of papermaking wastewater, mixing whey wastewater with the papermaking wastewater. The method also includes pre-treating the papermaking wastewater by acidification to stabilize the papermaking wastewater prior to entering an anaerobic reaction system.

Abstract: A nitrogen oxide gas sensor based on sulfur-doped graphene and a preparation method therefor. The method includes the following steps: 1) providing graphene and a micro heater platform substrate, and transferring the graphene onto the micro heater platform substrate; 2) putting the micro heater platform substrate covered with the graphene into a chemical vapor deposition reaction furnace; 3) performing gas feeding and exhausting treatment to the reaction furnace by using inert gas; 4) simultaneously feeding inert gas and hydrogen gas into the reaction furnace at a first temperature; 5) feeding inert gas, hydrogen gas and sulfur source gas into the reaction furnace at a second temperature for reaction to perform sulfur doping to the graphene; and 6) stopping feeding the sulfur source gas, and performing cooling in a hydrogen gas and insert gas shielding atmosphere.

Abstract: A semiconductor structure is disclosed. The semiconductor structure includes: a substrate of a first conductivity; a first region of the first conductivity formed in the substrate; a second region of the first conductivity formed in the first region, wherein the second region has a higher doping density than the first region; a source region of a second conductivity formed in the second region; a drain region of the second conductivity formed in the substrate; a pickup region of the first conductivity formed in the second region and adjacent to the source region; and a resist protective oxide (RPO) layer formed on a top surface of the second region. An associated fabricating method is also disclosed.