72results about How to "To achieve the purpose of classification" patented technology

The invention discloses an advertisement pushing method and device based on big data. The method comprises steps: an advertisement pushing need of an advertiser is received; according to the advertisement pushing need, user characteristics of an intention group with advertisement pushing are obtained; a user profile log database is queried and all users are classified according to the user characteristics of the intention group, wherein the user profile log database stores a correlation relationship between a user and a characteristic label of the user; and according to a classification result, a target user group is selected, and an advertisement is pushed to the target user group according to the advertisement pushing need. The user profile log database mode is adopted, through user characteristic matching with the intention group, the purpose of classifying the users is achieved, the technical effect of advertisement pushing accuracy is thus realized, and the technical problem of harassment of the advertisement to the user is further solved.

The invention provides a hyperspectral mixed pixel classification method based on neighbor cooperation enhancement, and the method comprises the steps: calculating a spectrum signature matrix of a plurality of target ground features through marked sample ground features; designing a multi-class classifier based on spectrum characteristics, and carrying out the classification of the ground features; carrying out the fusion of spatial structure features in a classification result, and extracting neighbor pixels; carrying out the class marking of unmarked hyperspectral ground features through the neighbor pixels; carrying out the classification and marking of the unmarked hyperspectral ground features through an interaction method; carrying out the further fusion of the spatial features of the target ground features in a mode of neighbor expansion, and completing the final classification and marking. According to the invention, the multi-class classifier is used for the simultaneous classification of ground features, and a problem that a conventional classification method cannot carry out the classification of background features is solved. Moreover, a mode of neighbor cooperation enhancement is employed for marking the unmarked ground objects step by step, thereby achieving the effective fusion of the spectrum features and spatial features of the ground features. The classification effect is good.

The invention relates to a vibration ore screening device which comprises a supporting frame, a screening motor, connection rods, first screening cloth, second screening cloth, third screening cloth, a first discharging plate, a second discharging plate, a third discharging plate and a feeding hopper. The feeding hopper is arranged on one side of the top of the supporting frame. The first screening cloth, the second screening cloth and the third screening cloth are arranged in the supporting frame evenly and parallelly from top to bottom at equal intervals, and the screening motor is connected with the first screening cloth, the second screening cloth and the third screening cloth through the connection rods. The first discharging plate, the second discharging plate and the third discharging plate are externally connected with one side, far away from the feeding hopper, on the outer side of the supporting frame evenly from top to bottom at equal intervals, and communicated with the first screening cloth, the second screening cloth and the third screening cloth respectively. The vibration ore screening device can perform multilevel classification on ores one time, so that manpower and material resources are saved, and production efficiency is increased greatly.

The invention discloses a slope mechanism applied to metal detection equipment. The structure of the slope mechanism applied to the metal detection equipment comprises a control maintenance cabinet, an adjusting controller, a metal detection frame, a conveying belt, a high-precision turnover metal classifying device, a supporting fixing frame and a working detection platform; the external side ofthe control maintenance cabinet is connected with the supporting fixing frame by using a bonding mode; the metal detection frame is mounted under the adjusting controller and connected with the adjusting controller by using a welding mode; the conveying belt is arranged under the metal detection frame; and the conveying belt is connected with the high-precision turnover metal classifying device. The slope mechanism applied to the metal detection equipment, disclosed by the invention, has the advantages that articles containing metal impurities can be accurately detected and classified to prevent the articles from being mixed with the metal impurities to affect the quality of the articles; moreover, when the equipment is used for performing metal detection and classification, a supporting and hydraulic shock absorbing effect on a classifying plate can be realized; an impact force on the classifying plate when the articles containing the metal impurities fall to the classifying plate isreduced; and the classifying plate is buffered and protected.

The invention discloses a data classification method and apparatus. The method comprises the steps of obtaining a style icon when an operation event is detected in any position of a user interface; in a process of dragging the style icon, determining at least one to-be-classified target object; and transferring the at least one to-be-classified target object to a directory in which the style icon is located. According to the method and the apparatus, the technical problem of low classification efficiency of target objects caused by the fact that a plurality of target objects displayed in the user interface are classified by performing multi-time dragging or clicking on the user interface in the prior art is solved.

The invention discloses a small sample multi-classification method based on a nested integrated depth support vector machine. According to the method, a two-layer integrated classification technologyis adopted to analyze multi-class classification tasks under the condition of small samples; for an internal layer, three different single-kernel functions (a radial basis kernel function, a Sigmod kernel function and a polynomial kernel function) are respectively given to a deep support vector machine (DSVM) of a sub-classifier, and a basis classifier is constructed through integrated classification; for an external layer, a'one-to-other 'strategy is used to convert a multi-class classification problem into a plurality of dichotomy problems. And finally, an AdaBoost framework is utilized to complete a multi-classification task. The method can effectively avoid an'over-fitting 'phenomenon under the condition of small samples, so that the accuracy of through-wall human body target recognition under the condition of small samples is improved.

The invention relates to an information processing method and device, computer equipment and a storage medium. The method comprises the following steps: acquiring original intelligence information, and performing quality identification on the original intelligence information to obtain intelligence quality of the original intelligence information; screening target intelligence information from the original intelligence information, wherein the target intelligence information is used for representing the intelligence information with the intelligence quality meeting the requirement; determining an intelligence label of the target intelligence information according to the information content of the target intelligence information, wherein the intelligence label of the target intelligence information comprises a first intelligence label, a second intelligence label and a third intelligence label; according to the intelligence label of the target intelligence information, performing clustering processing on the target intelligence information to obtain a plurality of intelligence information sets, wherein each intelligence information set comprises multiple pieces of target intelligence information; and storing the plurality of intelligence information sets in a preset database. By adopting the method, the utilization rate of intelligence information can be improved.

The embodiment of the invention provides a data processing method and device, a storage medium and an electronic device. The method comprises the steps:receiving the data information obtained by a sensor, the data information comprising image data and synchronous code information of the image data, wherein the synchronous code information comprises parameter information configured to the sensor by the coprocessor in real time and used for acquiring image data; and processing the image data according to the synchronous code information. According to the invention, the problem of image data processing in the prior art is solved, and the effect of reducing CPU performance overhead is achieved.

The invention relates to a composition for detecting human CYP2D6 gene polymorphism. The composition comprises primers for detecting the polymorphism of the C188T and/or G4268C locus of the CYP2D6 gene, and the primer for detecting C188T comprises a primer with the sequence as shown in SEQ ID NO. 1, SEQ ID NO. 2 and/or SEQ ID NO.3. The primer for detecting the G4268C locus comprises a primer withthe sequence as shown in SEQ ID NO. 4, SEQ ID NO. 5 and/or SEQ ID NO.6. The primer for detecting the C188T is combined with the primer for detecting G4268C, so as to detect the polymorphism of the C188T and the G4268C at the same time; the specificity is high; the sensitiveness is good; the repeatability is good.

The invention discloses a multifunctional safe sports equipment storage vehicle comprising a box body. The inner top wall of the box body is fixedly connected with an ultraviolet lamp; a first draweris arranged below the ultraviolet lamp and a containing box is fixedly connected to the upper surface of the first drawer; a groove is formed in the inner side wall of the lower box body, a second drawer is fixedly connected to one end of a second rolling wheel, a supporting frame is fixedly connected to the inner bottom wall of the second drawer, and sports equipment is arranged on the inner wallof the supporting frame. According to the multifunctional safe sports equipment storage vehicle, with the first drawer and the second drawer, the purpose of distinguishing and concluding is achieved;with arrangement of multiple grid plates, the sports equipment placed in the first drawer can be classified during use to achieve the purpose of classification; with the ultraviolet lamp, the purposeof enabling the sports equipment to be used safely is achieved; and because of the arrangement of the brush, the purpose of preventing the exterior of the sports equipment from being kept clean is achieved.

The invention relates to a garbage classification and disposal device, and belongs to the environment protection field, in particular to the garbage disposal field. By arranging the smart garbage classification device, there is no need to arrange garbage cans or other garbage storage facilities at homes or workplaces, and various kinds of garbage can be classified and disposed at any time. The purposes of classifying the garbage and disposing the garbage at any time are realized, and less indoor related facilities are needed.

The invention discloses a conveying system for cargo sorting. The conveying system for cargo sorting comprises a grabbing unit used for receiving a cargo grabbing command sent by a control unit and grabbing cargoes onto a transmission unit, a weighing unit used for acquiring the weight information of the cargoes and transmitting the weight information to the control unit, an image shooting unit used for acquiring the image information of the cargoes and transmitting the image information to the control unit, the transmission unit used for receiving an instruction sent by the control unit, andconveying the cargoes to a designated site, a label unit used for receiving an instruction sent by the control unit, and carrying out label printing, and the control unit used for sending the instruction and controlling the transmission unit to act, sending the cargo grabbing information and controlling the grabbing unit to carry out cargo grabbing, receiving the weight information and controllingthe grabbing unit to carry out weight classification on the cargoes, receiving the image information, processing the image information into cargo attribute information and carrying out attribute classification on the cargoes, and sending the instruction and controlling the label unit to carry out label printing. The purpose of classifying the cargoes is achieved, and the related personnel are facilitated to manage the cargoes.

The invention relates to a malicious traffic detection method based on image analysis, terminal equipment and a storage medium. The method comprises the following steps: S1, collecting to-be-detected network traffic data and malicious traffic data of a known type; s2, preprocessing all the data, and generating corresponding gray feature images; s3, for the gray feature image of each piece of data, performing noise reduction processing through singular value decomposition, extracting a low-frequency part of the image through discrete wavelet transform, and obtaining a binary hash code of the image through the low-frequency part of the image; and S4, according to the Hamming distance between the to-be-detected network traffic data and the binary hash code of the malicious traffic data of the known type, obtaining the similarity between the to-be-detected network traffic data and the binary hash code of the malicious traffic data of the known type, and when the similarity is smaller than a similarity threshold value, judging that the type of the to-be-detected network traffic data belongs to the type of the malicious traffic data of the known type. According to the method, the classification purpose can be realized without independently training samples and extracting features.

The invention relates to the technical field of sorting devices, in particular to a multi-view collecting module for red date sorting. According to the following scheme, the multi-view collecting module comprises a collecting box, wherein a through groove is formed in the top face of the collecting box, a filtering barrel is arranged inside the collecting box, a discharging port is formed in the side wall of the filtering barrel, a cover plate rotationally connected to the filtering barrel is arranged at the discharging port, a fourth rotating column, a first filtering plate, a bottom plate and a second filtering plate are arranged in the filtering barrel, the fourth rotating column is rotatably connected to the first filtering plate, the bottom plate and the second filtering plate, the first filtering plate is located between the bottom plate and the second filtering plate, the second filtering plate is located above the first filtering plate, the filtering barrel is divided into three cavities by the first filtering plate and the second filtering plate, rotating plates are arranged on the top surfaces of the first filtering plate and the second filtering plate, and the rotating plates are fixedly connected to the fourth rotating column. According to the module, the purpose of classifying the sizes of red dates is achieved, the red dates can be conveniently classified and discharged, and the working efficiency is improved.

The invention discloses a lung sound signal processing method for electronic auscultation, and belongs to the field of signal classification of machine learning. The problems are solved that since anexisting processing process of lung sound auscultation recording signals is cumbersome, the processing result has poor accuracy, and the final judgment result of the types of lung sounds is affected.The method includes the steps: performing band-pass filtering on the collected original lung sound signals, performing downsampling and normalization processing sequentially so as to obtain to-be-trained lung sound signals, and adopting multiple convolution units to process the to-be-trained lung sound signals so as to obtain lung sound signal feature vectors, wherein the connection mode of the multiple convolution units includes sequential connection and skip connection; and adopting a full-connection layer to process the lung sound signal feature vectors which are finally output by the convolution units so as to obtain classification results. The method can be applied to classification of the lung sound signals.

The invention discloses a multifunctional dressing change vehicle for nursing of burn departments, and relates to the technical field of nursing. The dressing change vehicle includes a box body; the left side surface of the box body is provided with a sliding groove; a sliding block is slidingly connected in the sliding groove; and the left side surface of the sliding block is fixedly connected toa material storage box. Through the mutual cooperation of a placing groove, clamping grooves, clamping plates, a lifting handle, a threaded column, a threaded cap, a connecting rod, a handle, a second rotating shaft and an adjusting disc, the purpose of classifying the medicine of each patient can be achieved; and through the rotation of the adjusting disc, the adjusting disc can drive the threaded column to rotate through the second rotating shaft, so that the threaded cap can drive the handle to move under the threaded connection with the threaded column, therefore, medical staff can adjustthe handle to the position that is convenient in holding, convenience can be brought for the work of working staff, and the practicability of the dressing change vehicle can be enhanced to a certainextent.