Abstract: A copy of a model comprising a plurality of trees is received, as is a copy of training set data comprising a plurality of training set examples. For each tree included in the plurality of trees, the training set data is used to determine which training set examples are classified as a given leaf. A blame forest is generated at least in part by mapping each training set item to the respective leaves at which it arrives.

Abstract: An optical fiber has corrugations on an outer surface to dissipate a portion of an input light beam, allowing the remaining portion of the light beam to be coupled to the optical fiber. By dissipating the portion of input light, damage to the optical fiber by uncoupled light is reduced.

Abstract: A nested anti-resonant nodeless hollow core fiber (NANF) enables transmission of multi-kilowatt, continuous wave (CW) light beams operating in wavelengths between 1050 nm and 1100 nm provided by single mode lasers. Such a NANF has little loss over kilometer ranges, and can be employed in long distance subsurface applications, such as in the petroleum industry.

Abstract: A fermented beverage and a preparation method are disclosed. The method uses ripened fruits and/or vegetables as raw materials. The method includes performing a natural ripening on a fruit and/or a vegetable at 15° C. to 30° C. for 5 days to 10 days to obtain a ripened fruit and/or vegetable; and inoculating lactic acid bacteria into the ripened fruit and/or vegetable, and performing a fermentation for 10 days to 30 days at 15° C. to 30° C. to obtain the fermented beverage. After 23 days of fermentation, the fermented beverage has a total flavonoid content up to 0.5 mg/mL to 2.0 mg/mL, which is beneficial to human health, a pH of 3.2 to 3.8 and a total titratable acidity of up to 12 mg/mL to 18 mg/mL. The fermented beverage requires a short fermentation cycle, which is beneficial to industrial production.

Abstract: An identification of an item that was misclassified by a classification model constructed in accordance with a machine learning technique is received. One example of such a machine learning technique is a random forest. A subset of training data, previously used to construct the model, and that is associated with the misclassified item is identified. At least a portion of the identified subset is provided as output.

Abstract: Described herein are advanced artificial intelligence agents for modeling physical interactions. In one embodiment, an apparatus to provide an active artificial intelligence (AI) agent includes at least one database to store physical interaction data and compute cluster coupled to the at least one database. The compute cluster automatically obtains physical interaction data from a data collection module without manual interaction, stores the physical interaction data in the at least one database, and automatically trains diverse sets of machine learning program units to simulate physical interactions with each individual program unit having a different model based on the applied physical interaction data.

Abstract: Methods and systems for budgeted and simplified training of deep neural networks (DNNs) are disclosed. In one example, a trainer is to train a DNN using a plurality of training sub-images derived from a down-sampled training image. A tester is to test the trained DNN using a plurality of testing sub-images derived from a down-sampled testing image. In another example, in a recurrent deep Q-network (RDQN) having a local attention mechanism located between a convolutional neural network (CNN) and a long-short time memory (LSTM), a plurality of feature maps are generated by the CNN from an input image. Hard-attention is applied by the local attention mechanism to the generated plurality of feature maps by selecting a subset of the generated feature maps. Soft attention is applied by the local attention mechanism to the selected subset of generated feature maps by providing weights to the selected subset of generated feature maps in obtaining weighted feature maps.

Abstract: The present invention belongs to the technical field of petrochemical engineering, and relates to a method for continuous post-treatment of benzotriazole (abbreviated as BTA) synthetic fluid. In particular, the present invention relates to a method for synthesizing BTA, including subjecting the BTA synthetic fluid to post-treatment steps of continuous acidification, water washing, extraction, back-extraction, dehydration, and distillation and the like. The method utilizes the difference in solubility of the BTA in water under different pHs to achieve separation by extraction without consuming a large amount of evaporation energy. The present invention is easy to operate, has little environmental pollution, high economic efficiency and low energy consumption, and is easily industrialized.

Abstract: Methods and systems for advanced and augmented training of deep neural networks (DNNs) using synthetic data and innovative generative networks. A method includes training a DNN using synthetic data, training a plurality of DNNs using context data, associating features of the DNNs trained using context data with features of the DNN trained with synthetic data, and generating an augmented DNN using the associated features.

Abstract: Exemplary etching methods may include flowing a fluorine-containing precursor into a substrate processing region of a semiconductor processing chamber. The methods may include flowing a hydrogen-containing precursor into the substrate processing region. The methods may include contacting a substrate housed in the substrate processing region with the fluorine-containing precursor and the hydrogen-containing precursor. The substrate may include a trench or recessed feature, and a spacer may be formed along a sidewall of the trench or feature. The spacer may include a plurality of layers including a first layer of a carbon-containing or nitrogen-containing material and a second layer of an oxygen-containing material. The methods may also include removing the oxygen-containing material.

Abstract: A method and apparatus for producing a high-pressure gas from an air separation unit is provided, in which the method includes the steps of introducing a cold air feed into a distillation column system under conditions effective for separating the cold air feed into a first air gas and a second air gas; withdrawing the first and second air gases from the distillation column system and warming said first and second air gases in a main heat exchanger, wherein the first air gas is withdrawn from the distillation column system at a medium pressure; splitting the first air gas into a first fraction and a second fraction; expanding the first fraction in a turbine; and compressing the second fraction in a booster to a pressure that is higher than the medium pressure, wherein the booster is powered by the turbine

Abstract: A flexible display module and a manufacturing method thereof are disclosed. The flexible display module includes a first flexible material layer and a second flexible material layer stacked on the first flexible layer, a sealing layer disposed between the first flexible material layer and the second flexible material layer, and a module layer disposed in the hollow of the sealing layer. The sealing layer has a hollow defined in a middle of the sealing layer. The area of the bottom area of the hollow is larger than the area of the module layer.

Abstract: An apparatus adapted for use in a plasma processing chamber is provided. An aluminum body with at least one surface is provided. An aluminum oxide containing aerosol deposition coating is disposed over the at least one surface of the aluminum body. An yttrium containing aerosol deposition coating is disposed over the aluminum oxide containing aerosol deposition coating.

Abstract: Textured silicon components of a semiconductor processing chamber having hillock-shaped or pyramid-shaped structures on its surface, and a method of texturing such silicon components. The silicon component can be selectively textured using chemical means to form the hillock-shaped structures to increase the surface area of the silicon component to improve polymer adhesion.

Abstract: A displaying device and a driving method thereof, wherein the displaying device includes a first display panel, a second display panel and a driving circuit; the first display panel includes a plurality of first pixel units that are able to display a color, wherein each of the first pixel units includes a plurality of sub-pixel units; and the second display panel is provided on a light exiting side of the first display panel, including a plurality of second pixel units; wherein a distribution density of the plurality of second pixel units is greater than a distribution density of the plurality of sub-pixel units.

Abstract: Methods and systems for budgeted and simplified training of deep neural networks (DNNs) are disclosed. In one example, a trainer is to train a DNN using a plurality of training sub-images derived from a down-sampled training image. A tester is to test the trained DNN using a plurality of testing sub-images derived from a down-sampled testing image. In another example, in a recurrent deep Q-network (RDQN) having a local attention mechanism located between a convolutional neural network (CNN) and a long-short time memory (LSTM), a plurality of feature maps are generated by the CNN from an input image. Hard-attention is applied by the local attention mechanism to the generated plurality of feature maps by selecting a subset of the generated feature maps. Soft attention is applied by the local attention mechanism to the selected subset of generated feature maps by providing weights to the selected subset of generated feature maps in obtaining weighted feature maps.

Abstract: An array substrate and a display panel. The array substrate includes a thin film transistor array layer including a driving transistor, a switching transistor, and a capacitor. The driving transistor includes a first active layer, a first gate insulating layer, a first gate, and an insulating dielectric layer sequentially stacked. The switching transistor includes a second active layer, a second gate insulating layer, and a second gate sequentially stacked. The insulating dielectric layer and the second gate insulating layer are located at a same layer. A thickness of the first gate insulating layer is greater than a thickness of the second gate insulating layer. The capacitor includes a first electrode plate and a second electrode plate. The first electrode plate and the first gate are disposed on same layer, and the second electrode plate and the second gate are disposed on same layer.

Abstract: A novel aromatic ring compound is represented by general formula (I). This compound, isomer, prodrug, solvate, pharmaceutically acceptable salt and pharmaceutical composition thereof are useful in preparing medicaments for treating depression and related symptoms.

Abstract: Skin smoothing is applied to images using a bilateral filter and aided by a skin map. In one example a method includes receiving an image having pixels at an original resolution. The image is buffered. The image is downscaled from the original resolution to a lower resolution. A bilateral filter is applied to pixels of the downscaled image. The filtered pixels of the downscaled image are blended with pixels of the image having the original resolution, and the blended image is produced.

Abstract: A component for use as part of a plasma processing chamber for processing a wafer is provided. The component comprises a component body of silicon carbide doped with at least one of tungsten, tantalum, or boron.