61 results about "Paired samples" patented technology

The invention relates to a high-speed high-temperature multifunctional friction and abrasion tester which is especially applicable to simulating the service conditions of the sealed pairs of turbojet engines and solves the problems that the friction pairs in the prior art can not meet the needs of high speed (over 100m/s) and high temperature (up to 1200DEG C) in the aspects of motion line speed and experiment temperature and can not simulate the high-speed high-temperature friction and abrasion conditions of the gas path sealing of the turbojet engines and the like. The tester is provided with a rotation main shaft, a rotation disc, a numerical control sliding table and heating devices, wherein the rotation disc is arranged on the rotation main shaft, a rotation sample is arranged on therotation disc, the numerical control sliding table is arranged at one side of the rotation main shaft, a translation sample is arranged on the numerical control sliding table and is opposite to the rotation sample to form a friction pair; and the heating devices are arranged at the side of the rotation sample and at the side of the translation sample. By adopting the invention, the high-speed stable rotation of the rotation main shaft of the tester and the rapid heating of surface of the friction pair sample can be realized.

The invention discloses a method for generating suggested keywords of sponsored search advertisement based on user feedback, which develops semantic relatedness between words and establishes a learning machine to evaluate the relatedness of word pairs by utilizing the feedback information of users on the relatedness of a small quantity of word pairs so as to select words which has high relatednesswith seed keywords describing advertiser products or service concepts as the suggested keywords to be recommended to advertiser users, wherein an active learning method is used for selecting a word pair sample with maximal information quantity to evaluate user relatedness so as to enhance the efficiency and the precision of generating related suggested keywords. The invention can effectively generate hundreds of to thousands of related words to be used as the suggested keywords to be recommended to the advertiser users corresponding to each seed keyword; in addition, the advertiser users carry out sponsored search bidding for the suggested keywords, thus the related click rate of the sponsored advertisement can be effectively enhanced, and the advertisement benefit is increased.

The invention discloses a vibration data fault classification method based on depth domain adaptation. The vibration data fault classification method comprises the steps of constructing a source domain containing a large amount of sample data and a target domain containing a small amount of sample data; constructing a classifier by using the sample data in the source domain; constructing paired samples according to the sample data in different domains in the source domain and the target domain; constructing a twin network, inputting the paired samples as training samples into the twin networkfor domain adaptation, and obtaining a final loss function of the paired samples; and optimizing the final loss function, and obtaining a trained fault classification model. According to the present invention, the problem that an existing deep network model is poor in diagnosis effect under the condition that fault sample data are insufficient is solved, a domain self-adaption method in deep learning and transfer learning is combined, the existing data are utilized to the maximum extent, the generalization capacity of the model is improved, and therefore better classification accuracy is obtained.

The invention relates to the field of biomedicine, in particular to a method and a device for detecting tumor mutation load based on a single sample. According to the device, a probe combination comprises probes that capture the exon region of a gene shown in Table 1, an intron promoter fusion breakpoint region of a gene shown in Table 2, and a coding region region of a gene shown in Table 3. A human tumor polygene detection kit comprises the probe combination. The TMB detection method based on the single sample comprises the following steps: access to target area of the tumor samples under test sequencing data, with the reference genome comparison, based on comparing the results obtained, detecting mutation loci, filtering original variable results with normal sample normal baseline database coincidence loci, filtering the high frequency reproductive mutation locus of a first cell mutation data set, screening the clonal somatic cell mutation locus of a second somatic cell mutation data set and calculating TMB. This method can accurately detect TMB of tumor samples without pairing samples.

The invention provides a facial expression conversion method based on identity and expression feature conversion, and mainly solves the problem of personalized facial expression. Most of the existingfacial expression synthesis work attempts to learn conversion between expression domains, so that paired samples and marked query images are needed. Identity information and expression feature information of an original image can be stored by establishing the two encoders, and target facial expression features are used as condition tags. The method mainly comprises the following steps: firstly, carrying out facial expression training, preprocessing a neutral expression picture and other facial expression pictures, then extracting identity characteristic parameters and target facial expressioncharacteristic parameters of a neutral expression, and establishing a matching model; Secondly, performing facial expression conversion, inputting the neutral expression picture into a conversion model, and applying model output parameters to expression synthesis to synthesize a target expression image. Pairing data sets of different expressions with the same identity are not limited any more, identity information of an original image can be effectively reserved due to existence of the two encoders, and conversion from a neutral expression to different expressions can be achieved.

The invention relates to an event related potential analyzing method based on a paired sample T test. The method includes: designing an electroencephalogram induction experiment containing two kinds of stimulations, using electroencephalogram collecting equipment to record multiple associated scalp electroencephalogram signals, and performing preliminary preprocessing; extracting ERP signals under the two kinds of stimulations; performing paired sample T test on the ERP signals under the two kinds of stimulations to determine a time period with significant difference; calculating the difference area of the ERP signals under the two kinds of stimulations in the time period with significant difference, and drawing a brain electrical activity mapping to determine the difference brain areas. The method has the advantages that the significant difference time period of the ERP under the two kinds of stimulations, the brain electrical activity mapping based on the ERP waveform difference area is drawn, and the different brain area distribution in the significant difference time period is obtained; the method is significant to ERP researches which are poor in signal to noise ratio and unobvious in single component, and a new idea is provided to the stripping of the ERP signals and spontaneous electroencephalogram.

The invention provides a method for simultaneously detecting a methylation level, genome variation and an insert fragment by utilizing a methylation non-bisulfite sequencing technology. The genome variation comprises gene mutation, copy number variation and structural variation. The invention also provides a device and equipment for implementing the method, and a corresponding computer readable medium. According to the invention, for methylated non-bisulfite sequencing data, one-step detection and analysis from offline data to multiple dimensions of methylation level, genome variation and insert fragments are realized, the method is suitable for whole genome and targeted capture data types of methylated non-bisulfite, and analysis of a single cancer sample and paired samples (cancer samples containing control samples) can be carried out.

A tumor mutation analysis method based on next-generation sequencing is characterized by comprising filtering a sample genome sequencing sequence; comparing the filtered sample genome sequencing sequence with a reference genome sequence; controlling the tumor sample comparison quality, wherein the type of the tumor sample is one of a tumor single sample or a tumor/control paired sample; performingsingle nucleic acid variation detection and insertion deletion marker detection according to the type of the tumor sample; and performing fusion detection according to the type of the tumor sample. The present invention provides an automated process for tumor variation biological information analysis, which can quickly and automatically detect variation markers such as SNV (single nucleic acid variation), indel (insertion deletion marker), fusion, CNV (gene copy number variation), TMB (tumor mutation load) and MSI, can obtain more information about the precise target of tumor treatment, and provides more help for patients to select targeted drugs with potential benefits.

One embodiment of the invention provides a single-sample-based next-generation sequencing tumor somatic cell variation detection device. The single-sample-based next-generation sequencing tumor somatic cell variation detection device includes a data acquisition module for acquiring sequencing data, the sequencing data comprising two blood samples of at least the same tumor and at least one tumor tissue sample; a data comparison module for comparing the sequencing data with a preset reference genome to obtain compared data; and a variation detection module for carrying out variation detection on the compared data and determining a somatic cell variation site and base variation of the tumor tissue sample. The technical scheme of the single-sample-based next-generation sequencing tumor somatic cell variation detection device carries out variation detection on the at least two blood samples so as to be combined into a variation background library, then filters reproductive variation by applying the variation background library in tumor tissue sample variation detection to obtain tumor-specific somatic variation, and detects the somatic variation of the tumor tissue sample under the condition that no paired sample exists, so that the false positive of tumor somatic variation detection is reduced.

The invention discloses a method for determining the reliability of a cigarette loose end detector that comes with a cigarette maker by hypothesis test and six sigma method, characterized by collecting an internally-discharged cigarette loose end rejecting amount and an externally-discharged cigarette loose end rejecting amount which are displayed by a cigarette maker industrial control display within the required time, simultaneously manually counting the amount of the internally-discharged cigarette loose ends and the amount of the externally-discharged cigarette loose ends in a cigarette loose waste tank, carrying out paired samples T test on the data; simultaneously according to the amounts of internally-discharged cigarette loose ends and externally-discharged cigarette loose ends which are not detected by the cigarette loose end detector, converting the internally-discharged cigarette loose end rate and externally-discharged cigarette loose end rate which are calculated by being detected by the cigarette loose end detector into a Sigma level. According to the invention, by means of hypothesis test and six sigma determination method, based on data analysis, the determination of the reliability and accuracy of the cigarette loose end detector that comes with the cigarette maker is realized accurately, objectively and effectively, thus the economic loss caused by shutdown detection is avoided.

The invention provides a method for predicting the wear resistance of lubricating base oil according to chemical structures. The method includes steps of 1), generating three-dimensional chemical structures of molecules of the lubricating base oil; 2), minimizing energy of the three-dimensional chemical structures; 3), computing an EVA (evaluation of an infrared vibration-based descriptor) parameter of each three-dimensional chemical structure; 4), preprocessing wear surface data of a friction pair sample after the lubricating base oil is applied to the friction pair sample; 5), performing regression on the EVA parameters and the wear surface data by means of partial least squares, and establishing a relationship between the preprocessed wear surface data and the EVA parameters so as to create a quantitative prediction model; 6), performing cross validation on the prediction model; 7), predicting the wear area of the friction pair sample according to the created prediction model when the lubricating base oil is applied to the friction pair sample in an experimental state. The method has the advantages that a computer-aided design process is introduced into the field of lubricating oil design for the first time on the basis of the quantitative structure-tribo-ability relationship, the method is beneficial to reducing lubricating oil design risks and research cost, and the lubricating oil development efficiency can be greatly improved.

The invention provides a high-speed impact film tribology testing machine with a novel structure. The high-speed impact film tribology testing machine comprises a high-speed rotary film sample stage composed of a high-precision gas suspension motor, a cantilever beam locating device, an impact friction pairing sample arranged at the front end of a cantilever beam, an impact movement system composed of a microscope for monitoring the distance between impact friction pairing sample and a film sample, and a signal acquisition system composed of a Doppler laser vibrometer and an acoustic vibration meter sensor. Compared with the prior art, the impact film tribology testing machine disclosed by the invention can precisely evaluate the performances such as impact frictional property, endurance life, contact force characteristics and the like of a micron/nanometer scale film under the condition of ultra-high speed movement and ultra-light impact load.

The invention provides a rough set and ANFIS (adaptive neuro-fuzzy inference system) based quantitative analysis method for the relation between glass batch and quality, belongs to the fields of automatic control, information technology and advanced manufacturing, and particularly relates to a rough set and ANFIS based quantitative analysis method used for improving quality of finished glass products. The method is characterized by comprising the following steps: firstly, generating glass batch ingredient-quality data pair sample library on the basis of historical production data, establishing a fuzzy information system based on fuzzy relation, simplifying the fuzzy information system on the basis of a rough set to determine inputting of a quantitative relation model between glass batch ingredients and finished product quality; and establishing the quantitative relation model between the glass batch ingredients and the finished product quality by adopting the ANFIS on the basis of analysis of a lag time constant for the glass batch ingredients and glass quality. The method can be used for analyzing the quantitative relation between the glass batch ingredients and the finished product quality, so that the batch ingredients are optimized, and the quality of the finished glass product is improved effectively.

The invention discloses a method for detecting copy number variation by means of a single sample based on second-generation sequencing technology, and a computer system. The method and the system can realize copy number variation (CNV) detection on the single sample according to sequencing original data without necessary or required factors in traditional methods such as comparison, sequencing depth, GC content correction and sample paring, thereby simplifying experiment and analysis steps, reducing cost, realizing high consistence between an analysis result and a traditional method, and performing effective correction on false positive and false negative results which are detected according to the traditional method through adding clinical verification results (such as FISH verification).

The microdialysis technique and device of the present invention can be used to study metabolic aspects of human and animal organs. When a microdialysis catheter is for example placed into the substance of a beating heart, there is always concern for problems with the catheter position. Further, there is always a risk for damage to heart tissue as well as to the catheter. The present invention is designed to avoid such damages, and a study has confirmed that data obtained from the epicardial probe of the present invention reflects the state of the myocardial metabolism. In anesthetized normoventilated pigs (n=20), a sternotomy was performed and a Gore-tex suture snare was placed around a branch of the left anterior descending (LAD) artery. One microdialysis probe was inserted in the myocardial tissue supplied by the snared LAD branch. Another microdialysis probe was placed on the epicardial surface above the myocardial probe. The protocol included four, ten minute ischemic periods (intervention group) followed by a forty minute ischemic period. The control group was only exposed to a forty minute ischemic period. It was possible to place the microdialysis probes of the present invention on the epicardial surface, and to recover samples reflecting the myocardial metabolic state of the heart. It was also possible to measure rapid changes in concentrations especially for lactate, where ischemia periods as short as ten minutes resulted in detectable increases. A similar pattern was found when analysing glucose and glycerol concentrations in paired samples. Data obtained from the device and method of the present invention, placed on the epicardial surface showed a similar pattern compared to data from a standard probe placed in the myocardium. Thus, the epicedial application of the present invention is useful for perioperative monitoring during cardiac surgery.

The invention discloses a method for establishing and providing a simulated data set required by CNV detection based on unknown CNV (Copy Number Variations) samples. According to the method provided by the invention, sequence data of normal samples are simulated and generated, the sample sequencing cost is reduced, the sequencing preference problem resulting from the deficiency of paired samples and unknown causes is solved, and the detection accuracy is improved.

The invention provides an asymmetric fine-grained infrared image generation system and method, and the system comprises an encoder which is used for obtaining an infrared image, carrying out the coding of the infrared image, and obtaining the representation z of a real sample x of the infrared image; The generator is used for obtaining the representation z of a real sample, sampling the distribution P (x | z, c) to generate a generated sample x 'of the infrared image, and carrying out paired sample matching on the real sample and the generated sample; The discriminator is used for matching themean value characteristics of the real samples with the mean value characteristics of the generated samples; The classifier is used for fitting posterior probability distribution P (c | x); And the infrared image generation model is used for generating a fine-grained condition image. According to the method, the problem of image generation failure caused by unstable generative adversarial model can be solved, generation of infrared image conditions can be controlled in a fine-grained manner, and the generated infrared image is good in diversity and authenticity.

The invention discloses a video pair mining method and device. The method comprises steps as follows: various video pair sample data containing a video identifier of a requested video, a video identifier of a clicked video and the click rate of the clicked video under the condition that a user watches the requested video are acquired, a probability matrix is determined according to the video identifier of the requested video, the video identifier of the clicked video and the click rate of the clicked video, an N-step transfer matrix corresponding to the probability matrix is calculated, and a video pair is mined according to the calculated N-step transfer matrix and the various acquired video pair sample data. With the method, the potential video pair can be effectively mined, that is, the video pair having the indirect request-click relation can be effectively mined, then the number of video pairs in a video recommendation system is increased, and the recommendation accuracy of the video recommendation system can be further increased.

The invention discloses a high-temperature high-speed pin disc friction testing machine. The testing machine is provided with a main transmission system, a loading system, a friction disc clamping device, a heating device, a numerical control feeding device and a data acquisition system; a friction disc sample is mounted on the friction disc clamping device; a friction pin sample is connected withthe numerical control feeding device through a cross beam and a lead screw; the friction disc sample and the friction pin sample are arranged opposite to each other to form a friction pair; a heatingand cooling device is arranged by the side of the friction disc sample and the friction pin sample; an input end of the data acquisition system is connected with a testing system; an output end of the data acquisition system is connected with a computer. By adopting the high-temperature high-speed pin disc friction testing machine, high-speed steady rotating of a rotating main shaft of the testing machine and quick heating of the surface of a friction pair sample can be realized.

The invention discloses a method and device for generating a binaural pin defect sample based on a generative adversarial network, and the method comprises the steps: collecting a binaural pin sample image, and obtaining a binaural pin sample image data set, wherein the binaural pin sample image data set comprises a normal sample image data set and a defect sample image data set; training a CycleGAN model by using the binaural pin sample image data set to obtain a paired sample image data set, wherein the paired sample images comprise a normal sample image and a correspondingly generated defect sample image; and training a Pix2Pix network model by using the paired sample image data set to obtain a binaural pin defect sample based on the generative adversarial network. According to the method, characteristics of various generative adversarial network models are combined, defect sample images with vivid corresponding effects can be effectively generated by processing normal samples, a balanced data set with rich characteristics is provided for subsequent training of a defect detection model, and thus the model obtains a good training effect.

The invention discloses a semantic question-answering method. The semantic question-answering method comprises the steps of creating an inverted sorting index library corresponding to preset questiondata in a question library; generating question pair sample data by utilizing the inverted sorting index library, and converting the question pair sample data into a character coding sequence; determining a loss function of a neural network model according to the character coding sequence and a part-of-speech coding sequence; determining a preset question vector corresponding to the preset question data by using the neural network model; determining a set of candidate question data similar to user question data in the question library according to the inverted sorting index library; calculating text vector similarity between the user question data and the candidate question data; and setting the candidate question data with the highest vector similarity as target question data, and outputting preset answer data corresponding to the target question data. According to the invention, the semantic question-answering accuracy can be improved. The invention further discloses a semantic question-answering device, electronic equipment and a storage medium, and the device has the above beneficial effects.

The invention discloses a plasma metabonomics analysis method of patient in primary osteoporosis based on somnipathy. The method comprises the following steps: sample preparation: collecting plasma sample of the primary osteoporosis patient of somnipathy people in sleep cognitive behavioral therapy, treating the plasm sample, and then performing the ultra-high-performance liquid chromatography-mass spectrometric detection analysis; determining the somnipathy according to a Pittsburgh sleep quality index table, wherein the ultra-high performance liquid chromatography-mass spectrometric detection analysis is as follows: performing chromatographic separation on a serum sample, and adopting mass spectrometric detection analysis; performing data analysis extraction and aligning instrument original data to acquire the data which is free from noise interference and can be used for statistical analysis; analyzing the data through the orthogonal partial least square judgement analysis and paired sample T test to acquire the difference metabolite between two groups, and performing initial identification and verification on the substances by combining the existing literature report. The method can provide beneficial technical support for the early prediction and prognosis of the somnipathy people in the primary osteoporosis occurrence.

The invention relates to a training method and device for a picture quality score model. The training method comprises the following steps: obtaining a plurality of original samples, wherein the original samples comprise sample pictures and quality scores corresponding to the sample pictures; determining a plurality of pairing samples according to the quality scores corresponding to the sample pictures, wherein the pairing samples comprise paired sample pictures and quality comparison results corresponding to the paired sample pictures; training a double-path neural network based on the pairedsamples, and determining a single-path neural network in the double-path neural network as a feature extraction network, the feature extraction network being used for extracting feature vectors of the sample pictures; and training a preset machine learning model based on the feature vector of the sample picture to obtain the picture quality score model. The training method can train a picture quality score model with high accuracy based on a small number of samples.

The embodiment of the invention discloses a data processing method and device and a computer readable storage medium. According to the embodiment of the invention, the method comprises: collecting positive word pair sample data and negative word pair sample data; training the auto-encoder according to the positive word pair sample data and the negative word pair sample data to obtain a trained auto-encoder; extracting feature information corresponding to the positive word pair sample data and the negative word pair sample data through the trained auto-encoder; inputting the feature informationinto a binary classifier for training to obtain a trained binary classifier; and combining the trained auto-encoder and the trained binary classifier to identify the hyponymy relationship of the to-be-identified word pair data. Therefore, the positive and negative word pair sample data simultaneously train the auto-encoder, and the feature information corresponding to the positive and negative word pair sample data is extracted based on the trained auto-encoder to perform combined training on the binary classifier, thereby realizing accurate identification of the hyponymy relationship, and greatly improving the data processing efficiency and the judgment accuracy of the hyponymy relationship.

The embodiment of the invention discloses a method and device for generating a model. One specific embodiment of the method comprises the steps of obtaining a universal corpus pair sample set in response to a received corpus model generation request, and obtaining a basic model by utilizing the universal corpus pair sample set, obtaining a domain corpus pair sample set, and iteratively executing abasic model parameter updating step comprising substeps of respectively obtaining a first preset number of general corpus pair samples and a second preset number of domain corpus pair samples from the obtained general corpus pair sample set and the domain corpus pair sample set, training model parameters of a basic model by sequentially utilizing the obtained universal corpus pair samples and domain corpus pair samples, and updating the basic model, and in response to the determined basic model parameter updating step, iteratively executing K times, and determining the basic model updated bythe Kth iteration as a corpus model. According to the embodiment of the invention, the corpus model with better reliability is obtained by training the sample with fewer domain corpora.

The invention discloses a leaf area correction coefficient acquisition method for Tinospora crispa. The leaf area correction coefficient acquisition method comprises the steps of S100, collecting leaves of Tinospora crispa as sample leaves; S200, measuring the length and the width of each sample leaf by adopting a leaf area analyzer measuring method, and measuring the length and the width of eachsample leaf by adopting a copying weighing method; S300, measuring the leaf area of each sample leaf by adopting the leaf area analyzer measuring method, and measuring the leaf area of each sample leaf by adopting the copying and weighing method; and S410, grading the sample leaves according to large, medium and small areas, and taking the leaf area of each sample leaf measured by the leaf area analyzer measuring method as a reference. According to the method, two leaf area measuring methods are adopted, the paired sample T test is performed on the obtained leaf areas according to the leaf size grade, two-factor (method and leaf size grading) variance analysis is performed on the length-width method correction coefficient calculated according to a scanning calculation method and a copyingweighing method, and the accuracy of the obtained correction coefficient is high.