Abstract: The neural network includes layers, and the layers each include a plurality of units, the plurality of units each have a weight coefficient associated with each input to a unit, and an importance parameter indicating importance of the weight coefficient. The method includes: in learning of a task, adjusting a first weight coefficient based on the importance parameter of the first weight coefficient and a first gradient of the first weight coefficient determined using a training set; and after the learning has been completed, determining, based on the training set, respective second gradients of a plurality of second weight coefficients included in a first layer including the first weight coefficient, and calculating the importance parameter of the first weight coefficient to be used in learning of a next task based on the respective second gradients.

Abstract: A symmetrical tri-axial parallel spindle head capable of multi-directional fixed-point rotation, comprising three branch chains (2), a moving base (5), and a fixed base (1). A machining spindle (6) is fixedly connected onto the moving base (5). Each of the branch chains (2) comprises a servo motor (22), a branch chain rod (21), a lead screw (23), and a guide nut (25). The lead screw (23) and a linear guide rail (24) are provided in an inner cavity on one end of the branch chain rod (21). One end of the lead screw (23) is connected to the servo motor (22), the lead screw (23) passes through the guide nut (25), and the other end of the lead screw (22) is connected to the branch chain rod (21) by means a rotating pair. The guide nut (25) is mounted on the linear guide rail (24) and is rotatably connected to the fixed base (1) through a first support disc (26). The other end of the branch chain rod (21) is rotatably connected to the moving base (5) through a second support disc (28).

Abstract: A symmetrical tri-axial parallel spindle head capable of multi-directional fixed-point rotation, comprising three branch chains (2), a moving base (5), and a fixed base (1). A machining spindle (6) is fixedly connected onto the moving base (5). Each of the branch chains (2) comprises a servo motor (22), a branch chain rod (21), a lead screw (23), and a guide nut (25). The lead screw (23) and a linear guide rail (24) are provided in an inner cavity on one end of the branch chain rod (21). One end of the lead screw (23) is connected to the servo motor (22), the lead screw (23) passes through the guide nut (25), and the other end of the lead screw (22) is connected to the branch chain rod (21) by means a rotating pair. The guide nut (25) is mounted on the linear guide rail (24) and is rotatably connected to the fixed base (1) through a first support disc (26). The other end of the branch chain rod (21) is rotatably connected to the moving base (5) through a second support disc (28).

Abstract: Techniques of prioritizing campaigns based on scrubbing rules using a bitmap are provided. A plurality of user rules is stored and a first set of users is identified for each user rule. A user rule index is generated based on the first set of users. A plurality of campaigns is stored and a second set of users is identified for each campaign. A campaign index is generated based on the second set of users. For each user indicated in the campaign index, a set of zero or more user rules is identified and a set of one or more campaigns is identified. A second set of campaign rules of a campaign that is in the set of one or more campaigns is identified. A determination is made whether at least one rule in the second set of campaign rules is in the first set of zero or more user rules.