Abstract: An article of footwear includes an upper, a sole structure coupled to the upper, at least one receiver integrated into an exterior surface of the upper, and a removable shield configured to selectively attach to the at least one receiver. The removable shield has an outward-facing ornamental exterior surface when attached to the at least one receiver.

Abstract: Fluid systems, foot support systems (including sole structures), and/or articles of footwear include: (a) a foot support bladder; (b) a fluid reservoir; (c) a pump (e.g., a foot-activated pump) including an inlet and an outlet, the outlet supplying fluid to one of the fluid reservoir or the foot support bladder; (d) a fluid line placing the foot support bladder or the fluid reservoir in fluid communication with the inlet of the pump; and (e) another fluid line placing the foot support bladder in fluid communication with the fluid reservoir. Such fluid systems, foot support systems, and/or articles of footwear further may include an electronic control system (e.g., an electronically controlled switch or valve), e.g., for changing at least one of the fluid lines between an open configuration and a closed configuration. The electronic control system may receive user input for adjusting foot support pressure (e.g., from a mobile electronic device).

Abstract: A method for assessing test adequacy of deep neural networks based on element decomposition is provided. The network testing is divided into black box testing and white box testing, of which key elements are decomposed and defined. Network parameters including a weight matrix and a bias vector are extracted. Importance values of neurons in individual layers of the deep neural network are calculated and clustered, and an importance value hot map of neurons in each layer is generated based on clustering results. Mutation testing, and index calculation and evaluation are performed.

Abstract: A method for rapidly acquiring multi-field response of mining-induced coal rock is provided. Based on 3D printing, the method achieves the control of material deformation and turns from 3D to 4D, thus saving manpower and material resources; further, repeated experiments may still be carried out under approximately the same conditions after each printing, so that the similarity of similarity simulations is greatly improved and a stable and reliable scientific law is conveniently obtained; besides, a BP neural network model may be built based on the data collected from similarity simulations to rapidly acquire an accurate and real coal seam mining response.