Abstract: Methods, systems, and non-transitory computer readable storage media are disclosed for determining parameters for digital campaign content in connection with executing digital campaigns using a rank-one assumption and a best-arm identification algorithm. For example, the disclosed system alternately explores response data in the first dimension and response data in the second dimension using the rank-one assumption and the best-arm identification algorithm to estimate highest sampling values from each dimension. In one or more embodiments, the disclosed system uses the estimated highest sampling values from the first and second dimension to determine a combination with a highest sampling value in a parameter matrix constructed based on the first dimension and the second dimension, and then executes the digital campaign using the determined combination.

Abstract: A method for recommending digital content includes: determining user preferences and a time horizon of a given user; determining a group for the given user based on the determined user preferences; determining a number of users of the determined group and a similarity of the users; applying information including the number of users, the similarity, and the time horizon to a model selection classifier to select one of a personalized model of the user and a group model of the determined group; and running the selected model to determine digital content to recommend.

Abstract: An adjustable attachment structure for a swimming machine is disclosed. The attachment structure is configured to interact with a frame of a pool. A user may interact with a locking switch to transition the attachment structure between a locked and an unlocked state. In the unlocked state, the swimming machine may be rotated about an axis to direct a generated current in a number of directions.

Abstract: A machine-learning framework uses partial-click feedback to generate an optimal diverse set of items. An example method includes estimating a preference vector for a user based on diverse cascade statistics for the user, the diverse cascade statistics including previously observed responses and previously observed topic gains. The method also includes generating an ordered set of items from the item repository, the items in the ordered set having highest topic gain weighted by similarity with the preference vector, providing the ordered set for presentation to the user, and receiving feedback from the user on the ordered set. The method also includes, responsive to the feedback indicating a selected item, updating the diverse cascade statistics for observed items, wherein the updating results in penalizing the topic gain for items of the observed items that are not the selected item and promoting the topic gain for the selected item.

Abstract: Methods, systems, and non-transitory computer readable storage media are disclosed for utilizing offline models to warm start online bandit learner models. For example, the disclosed system can determine relevant offline models for an environment based on reward estimate differences between the offline models and the online model. The disclosed system can then utilize the relevant offline models (if any) to select an arm for the environment. The disclosed system can update the online model based on observed rewards for the selected arm. Additionally, the disclosed system can also use entropy reduction of arms to determine the utility of the arms in differentiating relevant and irrelevant offline models. For example, the disclosed system can select an arm based on a combination of the entropy reduction of the arm and the reward estimate for the arm and use the observed reward to update an observation history.

Abstract: A child care apparatus includes a standing frame including a mounting portion, and a child support frame held on the standing frame via at least one hanging bar that has an end portion pivotally connected with the mounting portion, the hanging bar being rotatable for swinging the child support frame, wherein the end portion of the hanging bar is pivotally connected with the mounting portion at a position that is adjustable on the mounting portion.

Abstract: A wheeled carrying apparatus includes a standing frame, a handle frame, a wheel mount, a locking assembly and a release mechanism. The handle frame is rotatable between a first and a second position of different inclinations. The wheel mount carries a wheel and is rotatable relative to the standing frame for changing an orientation of the wheel axis. The locking assembly includes a wheel mount latch adapted to engage and rotationally lock the wheel mount relative to the standing frame, and to disengage and unlock the wheel mount. The release mechanism includes an operating part carried with the handle frame and is configured so that the operating part is operable to cause the wheel mount latch to disengage the wheel mount when the handle frame is in the first position, the operating part being operatively decoupled from the wheel mount latch when the handle frame is in the second position.

Abstract: A wheeled carrying apparatus includes a handle frame rotatable relative to a standing frame between a first and a second position, a wheel mount pivotally connected with the standing frame, a wheel mount latch adapted to engage and lock the wheel mount with respect to the standing frame and to disengage and unlock the wheel mount so that it is rotatable for changing orientation of a wheel axis, a link actuator coupled to the wheel mount latch via a linking part that is movable as the handle frame rotates for causing the wheel mount latch to engage or disengage the wheel mount, and a release mechanism including a cable assembly that couples an operating part carried with the handle frame to the wheel mount latch, the operating part being operable to cause the wheel mount latch to disengage the wheel mount in the first position of the handle frame.

Abstract: A child safety seat includes a first and a second restraining device, and an alert device and a first and a second switch serially connected in a series circuit. Each of the first and second restraining device is deployable to engage with a corresponding structure provided in a vehicle, each of the first and second restraining devices being operable independently and having an idle position and a deployed position. The first and second switches are respectively coupled operatively to the first and second restraining device, and are respectively closed when the first restraining device is in the deployed position and the second restraining device in the idle position. The alert device is thereby activable to emit an alert signal when the first restraining device is in the deployed position and the second restraining device in the idle position.

Abstract: An unlock device includes at least one operating member, a first unlock module and a second unlock module. The at least one operating member is movably disposed on a frame, wherein the frame is equipped with a first lock mechanism and a second lock mechanism. The first unlock module is movably disposed on the frame and connected to the first lock mechanism. The second unlock module is movably disposed on the frame and connected to the second lock mechanism. The at least one operating member moves to drive the first unlock module to move to unlock the first lock mechanism or drive the second unlock module to move to unlock the second lock mechanism.

Abstract: A child safety seat includes a seat shell having two sidewalls respectively provided at a left and a right side of the seat shell for restricting sideways movement of a child sitting on the seat shell, the two sidewalls including a first and a second sidewall, a buffering part movably connected with the seat shell, the buffering part being movable between a stowed position retracted toward the first sidewall and a deployed position protruding sideways from the first sidewall, a spring for biasing the buffering part toward the stowed position, and a latching mechanism including a latch carried with the buffering part, the latch being movable between a locking state for locking the buffering part in the deployed position and an unlocking state for unlocking the buffering part so that the buffering part is rotatable relative to the seat shell.

Abstract: A buckle assembly includes a first buckle component, a second buckle component and an operating component. The first buckle component includes a locked portion. The second buckle component includes a locking portion. The locking portion engages with the locked portion along a lateral direction of the buckle assembly when the second buckle component is mated with the first buckle component along a mating direction. The operating component is partially embedded in and partially exposed out of the second buckle component. The operating component is slidable relative to the second buckle component. The operating component drives the locking portion to move away from the locked portion for disengaging the locking portion from the locked portion during a sliding movement of the operating component relative to the second buckle component. The present invention has advantage of saving labor and easy operation.

Abstract: A high and low-pressure integrated air pump includes a single housing including an air inlet and an air outlet. A high-pressure pump is disposed within the housing and in fluid communication with the air inlet, and uses a first outlet passage to discharge to the air outlet. A low-pressure pump is also disposed within the housing and in fluid communication with the air inlet, and uses a second outlet passage to discharge to the air outlet.

Abstract: Recommendation systems and techniques are described that employ change point detection to generate recommendations for digital content. In one example, a change point detection technique is employed by a recommendation system to identify when a change point has occurred at a respective time step of a series of time steps. Detection of this change point may then be used by the recommendation system to reset the statistical model to address this change as well as generate a subsequent recommendation configured for exploration of reward distributions of the items of digital marketing content.

Abstract: An inflatable product can be inflated to a low pressure by a lower-power but high-volume electric pump, then “firmed” or fully inflated by an auxiliary pump. The auxiliary pump may be manually powered, such that the overall cost and complexity of the electric and auxiliary pumps are still lower than a high-power electric pump. The low power high volume electric pump can be easily powered by small batteries that will also provide an increased number of inflations compared to traditional high-power pumps for a given battery capacity. Using small batteries allows the pump to be very compact in size. The auxiliary pump may work through an auxiliary air chamber within the primary air chamber and fluidly connected thereto by a check valve. The electric pump may be reversible to provide for both inflation and deflation.

Abstract: A wheeled carrying apparatus includes a frame structure including a standing frame, a wheel mount carrying a wheel and pivotally connected with the standing frame, the wheel being rotatable relative to the first wheel mount about a wheel axis, a locking assembly including a wheel mount latch movably connected with the standing frame, the wheel mount latch being engaged with the wheel mount to rotationally lock the first wheel mount with respect to the standing frame, and disengaged from the wheel mount to unlock the wheel mount for rotation of the wheel mount to change an orientation of the wheel axis, an actuating assembly coupled to the wheel mount latch and including one or more linking part, and a release mechanism including an operating part provided on the frame structure that is operable to cause the actuating assembly to pull the wheel mount latch to disengage from the wheel mount.

Abstract: The present disclosure is directed toward systems, methods, and computer readable media for training and utilizing an item-level importance sampling model to evaluate and execute digital content selection policies. For example, systems described herein include training and utilizing an item-level importance sampling model that accurately and efficiently predicts a performance value that indicates a probability that a target user will interact with ranked lists of digital content items provided in accordance with a target digital content selection policy. Specifically, systems described herein can perform an offline evaluation of a target policy in light of historical user interactions corresponding to a training digital content selection policy to determine item-level importance weights that account for differences in digital content item distributions between the training policy and the target policy.

Abstract: The present disclosure is directed toward systems, methods, and computer readable media for training and utilizing an item-level importance sampling model to evaluate and execute digital content selection policies. For example, systems described herein include training and utilizing an item-level importance sampling model that accurately and efficiently predicts a performance value that indicates a probability that a target user will interact with ranked lists of digital content items provided in accordance with a target digital content selection policy. Specifically, systems described herein can perform an offline evaluation of a target policy in light of historical user interactions corresponding to a training digital content selection policy to determine item-level importance weights that account for differences in digital content item distributions between the training policy and the target policy.

Abstract: Systems and techniques for multi-task equidistant embedding are described that process categorical feature data to explore feature interactions. A digital analytics system enforces an equidistant relationship among features within a category while extracting high-order feature interactions by punishing both positive correlations and negative correlations among low-dimensional representations of different features. By enforcing an equidistant embedding, information is retained and accuracy is increased while higher order feature interactions are determined. Further, the digital analytics system shares knowledge among different tasks by connecting a shared network representation common to multiple tasks with exclusive network representations specific to particular tasks.

Abstract: Disclosed herein is a content-adaptive digital media content adjustment system, method and architecture. Playback rate of a media content item comprising audio and video content can be increased by removing a portion of the content item's audio content and a corresponding portion of the content item's video content. An increased playback rate of a multimedia content item is accomplished by decreasing the length of the multimedia content item by removing portions of the multimedia content item's audio and video content.