CN115358327A - Ocean thermocline data visualization method, device, equipment and medium based on PCA-SVM - Google Patents
Ocean thermocline data visualization method, device, equipment and medium based on PCA-SVM Download PDFInfo
- Publication number
- CN115358327A CN115358327A CN202211022965.1A CN202211022965A CN115358327A CN 115358327 A CN115358327 A CN 115358327A CN 202211022965 A CN202211022965 A CN 202211022965A CN 115358327 A CN115358327 A CN 115358327A
- Authority
- CN
- China
- Prior art keywords
- data
- thermocline
- ocean
- ocean thermocline
- svm
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/25—Integrating or interfacing systems involving database management systems
- G06F16/258—Data format conversion from or to a database
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Physics (AREA)
- General Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Mathematical Optimization (AREA)
- Artificial Intelligence (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Analysis (AREA)
- Computational Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Molecular Biology (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Algebra (AREA)
- Testing Or Calibration Of Command Recording Devices (AREA)
Abstract
The invention provides a PCA-SVM-based ocean thermocline data visualization method, device, equipment and medium, wherein the method comprises the following steps: acquiring an initial ocean thermocline data set; carrying out dimensionality reduction processing on initial ocean thermocline data based on a principal component analysis method, carrying out filtering value processing on the dimensionality reduction data based on a filtering control method, carrying out interpolation fitting on the filtered data by establishing a BP neural network model, and finally carrying out recognition classification on the interpolated and fitted data at different depths of ocean thermoclines through the established SVM model to obtain ocean thermocline target data; and carrying out visual processing on the target data of the ocean thermocline to obtain a visual ocean thermocline. According to the method, a series of preprocessing including dimensionality reduction, filtering, interpolation supplement and classification training is carried out on the ocean thermocline data, so that the accuracy of the ocean thermocline data information is greatly improved, and the problems of insufficient display accuracy and data loss caused by insufficient accuracy of the data information in the visualization process are solved.
Description
Technical Field
The invention relates to the technical field of data visualization, in particular to a PCA-SVM-based ocean thermocline data visualization method, device, equipment and medium.
Background
China is in a high-speed development stage of marine information research and application, the efficient processing and utilization of marine data is the basis for promoting the exploration and development of marine resources, china has a vast sea area, and the reasonable development and utilization of marine resources have important significance on economic development and technological progress of China. At present, with the characteristics of more and more frequent marine data surveying, more and more extensive surveying sea areas and the like, marine information data collected by surveying ships, satellites and the like has the characteristics of strong space-time relevance, complex and large scale, diversified data formats and the like, the characteristics of difficult analysis, low analysis efficiency and the like of data analysis software caused by the diversified and complicated data formats are brought, even different data processing software needs to be developed aiming at marine information with different data formats, and therefore the method has important significance for marine data information processing by standardizing the data formats and establishing unified data standards.
The classical marine thermocline analysis and processing method only carries out partial derivation and gradient value processing on step data, and can obtain the data change trend, but because various errors are inevitable in the data surveying and collecting process, the errors cannot be eliminated by partial derivation and gradient processing, and only the data change trend can be judged and observed. This causes problems such as accuracy reduction and data loss for ocean data information processing.
In conclusion, the problems that the precision of ocean thermocline data processing is not high, so that the display precision is not enough in the data visualization process, and data is lost exist in the prior art.
Disclosure of Invention
In view of the above, it is necessary to provide a method and a device for visualizing ocean thermocline data based on a PCA-SVM, so as to solve the technical problems in the prior art that the accuracy of processing the ocean thermocline data is not high, so that the display accuracy is not sufficient in the data visualization process, and the data is lost.
In order to solve the technical problem, in one aspect, the invention provides a PCA-SVM-based ocean thermocline data visualization method, which includes:
acquiring an initial ocean thermocline data set;
performing dimensionality reduction processing on the initial ocean thermocline data based on a principal component analysis method to obtain a dimensionality reduction ocean thermocline data set;
filtering the dimensionality-reduced ocean thermocline data based on a filtering control method to obtain a filtered ocean thermocline data set;
acquiring a target BP neural network model, inputting the filtered ocean thermocline data set into the target BP neural network model for interpolation fitting to obtain a fitted ocean thermocline data set;
acquiring a target SVM model, inputting the fitted ocean thermocline data set into the target SVM model to perform recognition and classification of ocean thermocline at different depths to obtain ocean thermocline target data;
and carrying out visualization processing on the ocean thermocline target data to obtain a visualized ocean thermocline.
In some possible implementations, the initial ocean thermocline data set includes four dimensions and corresponding seawater temperature data for the four dimensions, where the four dimensions include longitude, latitude, depth, and time.
In some possible implementation manners, the performing, based on a principal component analysis method, a dimension reduction process on the initial ocean thermocline data set to obtain a dimension reduced ocean thermocline data set includes:
carrying out zero-valued processing on the initial ocean thermocline data set to obtain a centralized matrix;
calculating covariance according to the centralized matrix to obtain a covariance matrix;
obtaining a covariance coefficient calculation eigenvalue according to the covariance matrix, and calculating an eigenvector according to the eigenvalue to construct an eigenvector matrix;
screening the eigenvector matrix according to a preset eigenvalue screening threshold value to obtain a primary screening matrix;
and carrying out scaling processing on the primary screening matrix to obtain a scaling matrix, wherein the scaling matrix forms the dimensionality reduction ocean thermocline data set.
In some possible implementation manners, the filtering the dimensionality reduced marine thermocline data based on a filtering control method to obtain a filtered marine thermocline data set, including:
carrying out abnormal data filtering value processing on the dimensionality-reduced ocean thermocline data set to obtain a preliminary filtering value ocean thermocline data set;
and performing mean square error calculation on data which are not processed by the filtered values in the preliminary filtered value ocean thermocline data set, and performing exception marking on data which are not processed by the filtered values and have the difference between the data values and the average value larger than the preset multiplying power mean square error to obtain the filtered ocean thermocline data set.
In some possible implementations, determining the target BP neural network model includes:
acquiring a data sample set of an ocean thermocline to be fitted;
determining the number of input layers, hidden layers and output layers in the BP neural network, and establishing an initial BP neural network model;
determining initial weight values and initial threshold values of the input layer, the hidden layer and the output layer, and performing iterative training on the initial BP neural network model based on the ocean thermocline data sample set to be fitted;
and updating the initial weight values and the initial threshold values of the input layer, the hidden layer and the output layer based on a genetic algorithm until the preset training precision is reached to obtain the target BP neural network model.
In some possible implementations, determining the target SVM model includes:
obtaining an ocean thermocline data sample set to be classified, and selecting different proportions to divide the ocean thermocline data sample set to be classified into a training set and a testing set;
determining an optimization objective function, a model training function, a decision function and an SVM kernel function to establish an initial SVM model;
and respectively carrying out training test on the SVM model based on the training set and the test set until an optimization punishment factor and a kernel function coefficient are determined to obtain a target SVM model.
In some possible implementation manners, the visualizing the target ocean thermocline data set and importing a visualization model to realize the visualization of the thermocline data includes:
performing matrix transformation on the target ocean thermocline data set to obtain a one-dimensional json data format file;
adding header file information to the one-dimensional json data format file, wherein the header file information comprises a time latitude and longitude range, time, ocean depth and a display mode to obtain a target json file;
and importing the target json file into a visualization model to realize the visualization of the ocean thermocline data.
On the other hand, the invention also provides an ocean thermocline data visualization device based on the PCA-SVM, which is characterized by comprising the following components:
the data acquisition module is used for acquiring an initial ocean thermocline data set;
the dimensionality reduction processing module is used for carrying out dimensionality reduction processing on the initial ocean thermocline data based on a principal component analysis method to obtain a dimensionality reduction ocean thermocline data set;
the filtering value processing module is used for carrying out filtering value processing on the dimensionality reduction ocean thermocline data based on a filtering control method to obtain a filtering ocean thermocline data set;
the fitting processing module is used for acquiring a target BP neural network model, inputting the filtered ocean thermocline data set into the target BP neural network model for interpolation fitting to obtain a fitted ocean thermocline data set;
the classification processing module is used for acquiring a target SVM model, inputting the fitted ocean thermocline data set into the target SVM model to perform recognition and classification of ocean thermocline at different depths to obtain ocean thermocline target data;
and the visualization module is used for carrying out visualization processing on the ocean thermocline target data to obtain a visualized ocean thermocline.
On the other hand, the invention also provides electronic equipment which comprises a memory, a processor and a computer program which is stored on the memory and can run on the processor, wherein when the processor executes the program, the ocean thermocline data visualization method based on the PCA-SVM is realized.
Finally, the invention also provides a computer-readable storage medium, on which a computer program is stored, which, when executed by a processor, implements the PCA-SVM-based ocean thermocline data visualization method described in the above implementation.
The beneficial effects of adopting the above embodiment are: according to the ocean thermocline data visualization method based on the PCA-SVM, on one hand, a series of preprocessing steps of PCA dimension reduction, abnormal value filtration, interpolation supplement, step marking and SVM classification training are carried out on the initial ocean thermocline data, and therefore the accuracy of the ocean thermocline data information is greatly improved. On the other hand, the processed ocean thermocline data information is converted into a json data format through matrix transformation, and the json data is displayed through a visual model, so that the ocean thermocline data can be subjected to dimension reduction and feature scaling processing through the ocean thermocline data visualization method based on the PCA-SVM, the data visualization operation on the thermocline is more visual, and the problem of data loss caused by insufficient display precision due to low precision of the ocean thermocline data information is solved.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the description below are only some embodiments of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained based on these drawings without creative efforts.
FIG. 1 is a schematic flow chart diagram of an embodiment of a PCA-SVM-based ocean thermocline data visualization method provided by the invention;
FIG. 2 is a schematic flow chart diagram of an embodiment of PCA dimension reduction processing provided by the present invention;
FIG. 3 is a flowchart illustrating an embodiment of exception data filtering according to the present invention;
FIG. 4 is a schematic flow chart of an embodiment of a target BP neural network model construction provided by the present invention;
FIG. 5 is a schematic diagram of an embodiment of the BP neural network error delivery method and network hierarchy relationship provided by the present invention;
FIG. 6 is a schematic flow chart of an embodiment of a BP neural network model based on a genetic algorithm provided in the present invention;
FIG. 7 is a schematic flow chart of one embodiment of a target SVM model construction method provided by the present invention;
FIG. 8 is a schematic flow chart of an embodiment of the ocean thermocline data processing based on PCA-SVM provided by the present invention;
FIG. 9 is a schematic flow chart diagram illustrating an embodiment of ocean thermocline data visualization provided by the present invention;
FIG. 10 is a schematic structural diagram of an embodiment of an ocean thermocline data visualization device based on PCA-SVM in the ocean provided by the invention;
fig. 11 is a schematic structural diagram of an embodiment of an electronic device provided in the present invention.
Detailed Description
The technical solution in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. It should be apparent that the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments, which can be obtained by a person skilled in the art without making any creative effort based on the embodiments in the present invention, belong to the protection scope of the present invention.
It should be understood that the schematic drawings are not necessarily to scale. The flowcharts used in this invention illustrate operations performed in accordance with some embodiments of the present invention. It should be understood that the operations of the flow diagrams may be performed out of order, and that steps without logical context may be performed in reverse order or concurrently. One skilled in the art, under the direction of this summary, may add one or more other operations to, or remove one or more operations from, the flowchart.
Some of the block diagrams shown in the figures are functional entities and do not necessarily correspond to physically or logically separate entities. These functional entities may be implemented in the form of software, or in one or more hardware modules or integrated circuits, or in different networks and/or processor systems and/or microcontroller systems.
Reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the invention. The appearances of the phrase in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. It is explicitly and implicitly understood by one skilled in the art that the embodiments described herein may be combined with other embodiments.
Before the description of the embodiments, the related words are paraphrased:
PCA: PCA (principal component analysis), which is a mathematical dimension reduction method, converts a series of variables that may be linearly related into a set of new linearly uncorrelated variables, also called principal components, using orthogonal transformation, thereby characterizing the data in smaller dimensions using the new variables.
SVM: an SVM (support vector machine) is a support vector machine, is a machine learning method capable of training, and is a supervised learning model for analyzing data in classification and regression analysis and a related learning algorithm. Given a set of training instances, each of which is labeled as belonging to one or the other of two classes, the SVM training algorithm creates a model that assigns the new instance to one of the two classes, making it a non-probabilistic binary linear classifier. The SVM model represents instances as points in space such that the mapping is such that the instances of the individual classes are separated by as wide a distinct interval as possible. The new instances are then mapped to the same space and the categories are predicted based on which side of the interval they fall on.
A thermocline: the Thermocline (Thermocline) is a thin layer which is located about 100-200 m below the sea surface and has great changes in temperature and density, and is a layer with a sharp drop in water temperature between an upper thin warm water layer and a lower thick cold water layer.
Based on the description of the technical terms, at present, with the characteristics of more and more frequent marine data exploration, more and more extensive exploration sea areas and the like, marine information data collected by exploration ships, satellites and the like have the characteristics of strong space-time relevance, complex and huge scale, diversified data formats and the like, and the effective analysis and processing of marine thermocline data information has important significance on marine exploration and submarine battle. In the prior art, the marine thermocline analysis and processing method only carries out partial derivation and gradient value processing on step data, and although the data change trend can be obtained, various errors are inevitable in the data surveying and collecting process, so that the errors cannot be eliminated by partial derivation and gradient processing, and only the data change trend can be judged and observed. This causes problems such as accuracy reduction and data loss for ocean data information processing. Therefore, the invention aims to provide a PCA-SVM-based ocean thermocline data visualization method, which can improve the accuracy of ocean thermocline data processing and solve the problem of data loss caused by insufficient display accuracy due to low accuracy of ocean data information, so that the ocean thermocline data is more reliable, accurate and visual from collection, analysis and processing to display.
Specific examples are described in detail below:
the embodiment of the invention provides a PCA-SVM-based ocean thermocline data visualization method and device, electronic equipment and a storage medium, which are respectively described below.
As shown in fig. 1, fig. 1 is a schematic flow chart of an embodiment of a PCA-SVM-based ocean thermocline data visualization method provided by the present invention, and as shown in fig. 1, the PCA-SVM-based ocean thermocline data visualization method includes:
s101, acquiring an initial ocean thermocline data set;
s102, performing dimensionality reduction processing on the initial ocean thermocline data based on a principal component analysis method to obtain a dimensionality reduction ocean thermocline data set;
s103, filtering value processing is carried out on the dimensionality-reduced ocean thermocline data based on a filtering control method to obtain a filtered ocean thermocline data set;
s104, obtaining a target BP neural network model, inputting the filtered ocean thermocline data set into the target BP neural network model for interpolation fitting to obtain a fitted ocean thermocline data set;
s105, obtaining a target SVM model, inputting the fitting ocean thermocline data set into the target SVM model to perform recognition and classification of ocean thermocline at different depths to obtain ocean thermocline target data;
and S106, carrying out visualization processing on the ocean thermocline target data to obtain a visualized ocean thermocline.
Compared with the prior art, on one hand, the ocean thermocline data visualization method based on PCA-SVM provided by the embodiment of the invention greatly improves the accuracy of the ocean thermocline data information by performing a series of preprocessing of PCA dimension reduction, outlier filtering, interpolation supplement, step marking and SVM classification training on the initial ocean thermocline data. On the other hand, the processed data information of the ocean thermocline is converted into a json data format through matrix transformation, and the json data is displayed through a visualization model, so that the dimension reduction and characteristic scaling processing can be performed on the data of the ocean thermocline by the ocean thermocline data visualization method based on the PCA-SVM, the data visualization operation on the thermocline is more visual, and the problem of data loss caused by insufficient display precision due to low precision of the data information of the ocean thermocline is solved.
In some embodiments of the invention, the initial ocean thermocline data set is obtained by various types of survey vessels, marine detection sensors, satellites, and other detection equipment, wherein the ocean thermocline data comprises four dimensions and seawater temperature data corresponding to the four dimensions, wherein the four dimensions comprise longitude, latitude, depth, and time.
In some embodiments of the present invention, as shown in fig. 2, fig. 2 is a schematic flowchart of an embodiment of PCA dimension reduction processing provided by the present invention, and step S102 includes:
s201, carrying out zero-valued processing on the initial ocean thermocline data set to obtain a centralized matrix;
s202, calculating covariance according to the centralized matrix to obtain a covariance matrix;
s203, obtaining a covariance coefficient calculation eigenvalue according to the covariance matrix, and calculating an eigenvector according to the eigenvalue to construct an eigenvector matrix;
s204, screening the eigenvector matrix according to a preset eigenvalue screening threshold to obtain a primary screening matrix;
s205, carrying out scaling processing on the primary screening matrix to obtain a scaling matrix, wherein the scaling matrix forms the dimensionality reduction ocean thermocline data set.
In a specific embodiment of the present invention, the ocean temperature data includes four dimensions, which are longitude, latitude, depth, and time, and the four dimensions can determine the ocean temperature data at any time and any position around the world.
Firstly, zero equalization processing is carried out on collected ocean thermocline data, each column of two-dimensional data at a certain moment forms a new matrix Y of N x M, and N corresponds to the row number of the data, namely the basic characteristic attribute of the data. Based on the m-dimensional data and the thermocline mark K, calculating the average value of all dimensions under the Y matrix, and subtracting the average value from the data under different dimensions, wherein the calculation method comprises the following steps:
wherein Y is a matrix formed by data of seawater temperature at a certain sampling moment in a certain latitude range, m is a matrix dimension, and Y is m Is the mean value of the m column of the matrix, Y m For the mth column of the Y matrix, N corresponds to the number of rows of the data matrix, Y in Is the raw data matrix which is not processed yet.
Then, introducing a covariance matrix, wherein different data characteristics in the Y matrix are N, the correlation among different data is judged according to a covariance coefficient, the covariance coefficient is positively correlated with the influence of the characteristics on the whole data set, and the specific calculation method is as follows:
wherein M is cov Covariance is solved in different dimensions for seawater temperature data to obtain a covariance matrix, covariance coefficients are used for eigenvalue calculation, after eigenvalues are calculated, a new matrix xi is constructed by using eigenvectors formed by the eigenvalues, and then the eigenvalues are screened to remove eigenvalues outside a threshold value. Setting lambda j The j-th row characteristic value is obtained by the following concrete implementation formula:
ξ=[ξ 1 ,ξ 2 ,L,ξ m ],|λ j |>ε
where xi is a new matrix composed of characteristic vectors of the data matrix, λ j And f, screening a threshold value for the characteristic value of the jth row, wherein epsilon is the set characteristic value.
Finally, after the data are primarily processed, a new vector space is formed, and the new vector space is as follows:
Y′=[Y 1 ,Y 2 ,L,Y m ],ξ j →Y j
and Y' is a new vector space formed after the data processing, and the new vector space actually completes primary screening. However, compared with other data formats, the thermocline data has larger coordinate magnitude difference under one-dimensional depth, so that the processed data sample is zoomed for more intuitive display on a coordinate axis. The calculation method is as follows:
wherein, delta n Representing the standard deviation of unprocessed data, and obtaining a data matrix Y 'for training after scaling processing' n And importing the data matrix into a transform function in a standarsscaler for scaling.
According to the embodiment of the invention, the dimensionality of the ocean thermocline data is reduced by using the PCA to perform dimensionality reduction on the ocean thermocline data, and the data subjected to dimensionality reduction is further subjected to scaling processing, so that the accuracy of the ocean thermocline data processing is initially improved.
Further, in some embodiments of the present invention, as shown in fig. 3, fig. 3 is a schematic flowchart of an embodiment of abnormal data filtering processing provided by the present invention, and step S103 includes:
s301, carrying out abnormal data filtering value processing on the dimensionality-reduced ocean thermocline data set to obtain a preliminary filtering value ocean thermocline data set;
s302, mean square deviation calculation is carried out on data which are not processed by the filtered values in the preliminary filtered value ocean thermocline data set, and abnormal marking is carried out on data which are not processed by the filtered values and have the difference between the data values and the average value larger than a preset multiplying power mean square deviation, so that the filtered ocean thermocline data set is obtained.
In the specific embodiment of the invention, because uncontrollable errors are inevitably generated by various instruments in the measurement process due to external factors in the marine surveying process, abnormal data still needs to be filtered for the data after the characteristic quantity is extracted. Errors of marine data information usually come from two aspects, the first is errors caused by position errors due to abnormal signals of a measuring ship and a GPS, and the errors are obvious in characteristics after data clustering analysis; secondly, due to errors generated by faults of the measuring device, the errors have more intuitive data characteristics and are obviously different from data under different dimensions.
In order to better analyze data, the invention adopts a filtering control method to filter abnormal data, and compares the verified data sets in a small range so as to remove the abnormal data.
Firstly, a certain observation element data set of the thermocline is set as G i (i =1,2.., N), observation element dataset G i The physical meaning is the depth of seawater in a certain latitude and longitude area at a certain moment, and the set value is G i At (G) min ,G max ) Within the range, there is no error. If the range is exceeded, the value is deleted and the average value of reasonable data in the observation range is usedFilling, wherein N is the number of the data sets, and the specific processing method is shown by the following formula:
in order to better check the effect of the abnormal data after filtering, the mean square error calculation is performed on the data without filtering, and if the difference between the data value of a certain point and the mean value is greater than n times the mean square error, the value of the data point is marked as abnormal, and the specific processing process is as follows:
wherein Var represents the mean square error in the normal data range, the value of n determines the accuracy of marking the abnormal data and the proportion of marked abnormal data, and the value of n is between 3 and 6, which has great influence on the subsequent data analysis.
According to the embodiment of the invention, abnormal data filtering processing is carried out on the data subjected to dimensionality reduction, mean square error calculation is carried out on the data which is not subjected to value filtering processing, abnormal values conforming to abnormal characteristics are subjected to abnormal marking, and the precision of processing the data of the ocean thermocline is further improved.
Further, in some embodiments of the present invention, as shown in fig. 4, fig. 4 is a schematic flowchart of an embodiment of constructing a target BP neural network model provided by the present invention, and the process of obtaining the target BP neural network model in step S103 includes:
s401, obtaining a sample set of ocean thermocline data to be fitted;
s402, determining the number of input layers, hidden layers and output layers in the BP neural network, and establishing an initial BP neural network model;
s403, determining initial weight values and initial threshold values of the input layer, the hidden layer and the output layer, and performing iterative training on the initial BP neural network model based on the ocean thermocline data sample set to be fitted;
s404, updating initial weight values and initial threshold values of the input layer, the hidden layer and the output layer based on a genetic algorithm until preset training precision is achieved, and obtaining the target BP neural network model.
In a specific embodiment of the invention, after the filtering and marking of the thermocline data anomaly are completed, in order to keep the continuity of the ocean data information under unit precision, interpolation fitting needs to be performed on the processed data. The filtered data of the temperature jump layer is subjected to interpolation fitting by adopting a Back Propagation (BP) neural network, so that the continuity of the data is ensured, and preparation is made for subsequent data visualization.
The BP neural network has wide application in data processing, is suitable for interpolation fitting processing of a thermocline, and has the implementation process that: the method comprises the steps of repeatedly training a large number of sample data sets needing fitting, reversely transmitting network errors in a training model to adjust a threshold value and a weight coefficient in a whole system network, carrying out gradient search on thermocline data with gradient change to enable an error function to show negative gradient reduction, enabling an output new data set to be close to an expected data set, and achieving interpolation fitting.
Fig. 5 is a schematic diagram of an embodiment of the error transmission manner and the network hierarchy relationship of the BP neural network provided by the present invention.
The data acquisition graduation of the ocean thermocline determines the number of output layers and input layers in the neural network, the setting of the hidden layers is often obtained through repeated tests, too many neurons in the hidden layers can influence the calculation speed of the neural network system, the overfitting problem can be caused, and unnecessary high-precision data can be generated. The quality of data fitting can be directly influenced by the fact that the number of the hidden layer neuron networks is not enough. Therefore, the invention adopts the following relation to select the number of the hidden layer neurons:
where m represents the hidden layer neuron number, k represents the input layer neuron number, l represents the output layer neuron number, and b is a constant between the intervals [1,15 ].
The system stimulus function is defined as:
the training algorithm uses the L-M (Levenberg-Marquardt) algorithm. For the data of the ocean temperature jump layer, the L-M algorithm has certain advantages, and if the gradient of the data in a certain region is reduced too fast, the L-M algorithm adjusts the value of tau so that the training algorithm is changed into a Gaussian-Newton algorithm; if the gradient is descending too slowly, the L-M algorithm adjusts the value of lambda to change the training algorithm into a gradient descending method, and it can be seen that lambda is mainly used for calculating the step increment of the system. The specific calculation method is as follows:
wherein G is k For observing the element data set matrix G i After being filteredG is G k Of the Jacobian matrix, Δ T Step increment of the algorithm, xi, is mentioned in step one, and is a matrix formed by characteristic vectors, tau is an algorithm adjusting threshold, the predicted output of the neural network system is recorded as U (n), the actual output is V (n) after model training, and the input and output are expressed as follows:
U(N)=[u 1 ,u 2 ,L,u N ]
V(N)=[v 1 ,v 2 ,L,v N ]
the mean square error is selected to evaluate the error between the actual output and the predicted output, namely, the mean square error is used to evaluate the data interpolation fitting effect, the MSE is a mean square error evaluation coefficient, the N is the dimension of input data, u is the dimension of the input data i For the ith column of the output prediction matrix, v i For actually outputting the ith column of the matrix, the specific implementation method is as follows:
in the actual processing process, the weight coefficients of the neural network model converge to the local extreme points in the process of training using the sample data of the temperature jump layer. Therefore, in order to achieve a better neural network fitting effect, the Genetic Algorithm (Genetic Algorithm) is used for global neural network optimization when the iteration weights and the threshold values are defined, and the implementation process is shown in fig. 6, where fig. 6 is a schematic flow diagram of an embodiment of the BP neural network model based on the Genetic Algorithm provided by the present invention.
According to the embodiment of the invention, the BP neural network model based on the genetic algorithm is designed, and interpolation fitting is carried out on the missing part in the ocean thermocline data after the abnormal data processing and marking, so that the integrity of the ocean thermocline data set is ensured, the precision of the ocean thermocline data is further improved, and reliable guarantee is provided for visualization of subsequent data.
Further, in some embodiments of the present invention, as shown in fig. 7, fig. 7 is a flowchart illustrating an embodiment of constructing a target SVM model provided by the present invention, and the process of obtaining the target SVM model in step S104 includes:
s701, obtaining an ocean thermocline data sample set to be classified, and selecting different proportions to divide the ocean thermocline data sample set to be classified into a training set and a testing set;
s702, determining an optimization objective function, a model training function, a decision function and an SVM kernel function to establish an initial SVM model;
and S703, respectively carrying out training test on the SVM model based on the training set and the test set until an optimization punishment factor and a kernel function coefficient are determined, and obtaining a target SVM model.
In the specific embodiment of the invention, the data after the dimensionality reduction processing, the abnormal data filtering and the interpolation fitting are used as the input characteristics of a Support Vector Machine (SVM), the classification of ocean temperatures at different depths is realized, an ocean thermocline is recognized, and the data recognition and processing efficiency of the thermocline is greatly improved by the trained SVM model.
To complete training of the SVM model, a large amount of data processed in the above embodiments needs to be input into an SVM algorithm, a data set formed by the processed data is called a training set, the training set directly influences the values of a penalty factor C and a kernel function coefficient in the SVM model, and a part of unprocessed data in the test set is used for checking the model effect. The data type of the training set is a two-dimensional matrix formed by floating point type data, the trained SVM model inputs the unprocessed marine information two-dimensional data matrix, and outputs the analyzed marine information two-dimensional data matrix which can be used for prediction and display.
A Support Vector Machine (SVM) is a linear classification model for carrying out binary classification on ocean information data according to a supervised learning mode. The SVM usually uses the maximum margin plane solved by the preprocessed sample data as the decision boundary to obtain the optimized objective function as shown below.
Wherein m represents the process of dimension reductionIn the method, the dimension of a new matrix formed by re-dividing the ocean information data matrix is obtained, j represents the number of sample indexes, y j Sample data representing training, x i The number of lines of ocean information data samples is represented, omega and b are target decision parameters, C is a penalty factor in SVM, and delta i Is the relaxation variable. Considering the effect and speed of the SVM model algorithm, model training selects a Gaussian kernel function as follows:
K(y,c)=exp(-γ||y-y′|| 2 )
two core indexes optimized by the SVM model are a penalty factor C and a kernel function coefficient gamma respectively, the penalty factor C mainly adjusts the step size and the accuracy of the classifier, the lambda determines the plane space of ocean information data mapping influenced by the kernel function, the kernel function coefficient and the penalty factor have large influence on model training classification, and the kernel function coefficient and the penalty factor directly determine whether the SVM model can be applied to ocean information data samples, so that the penalty factor C and the kernel function coefficient lambda need to be optimized and determined. Based on the above, a decision function and an SVM kernel function are obtained, which are respectively as follows:
and processing the ocean thermocline data subjected to PCA dimension reduction, abnormal data filtering and interpolation fitting based on the established and trained SVM model to obtain ocean thermocline target data, wherein the ocean data can be used as a predicted and displayed ocean information two-dimensional data matrix.
To more clearly illustrate the data processing flow in the embodiment of the present invention, please refer to fig. 8, and fig. 8 is a schematic flow diagram of an embodiment of the PCA-SVM based ocean thermocline data processing provided by the present invention.
According to the embodiment of the invention, the classification of the ocean temperatures at different depths is realized by establishing the SVM model, the ocean thermocline is identified, the accuracy of the data of the ocean thermocline is further improved, and the visual display accuracy of the data of the ocean thermocline is guaranteed.
Further, in some embodiments of the present invention, as shown in fig. 9, fig. 9 is a schematic flowchart of an embodiment of visualizing marine thermocline data provided by the present invention, and step S107 includes:
s901, performing matrix transformation on the target ocean thermocline data set to obtain a one-dimensional json data format file;
s902, adding header file information to the one-dimensional json data format file, wherein the header file information comprises a time longitude and latitude range, time, ocean depth and a display mode to obtain a target json file;
and S903, importing the target json file into a visualization model to realize visualization of the ocean thermocline data.
In a specific embodiment of the present invention, the data display method, through the steps in the data processing embodiment, the processed data is still two-dimensional data, but after filtering, interpolation fitting, and SVM model processing, the data accuracy and reliability are greatly improved, and within a certain latitude and longitude range, the data accuracy may be from 0.5 ° × 0.5 ° to even 0.25 ° × 0.25 °. Firstly, matrix transformation is carried out on a two-dimensional data matrix by using MATLAB to obtain a one-dimensional json data format, a header file is added to the json file, header file information comprises a time longitude and latitude range, time, ocean depth and a display mode, and then the processed json file is imported into a visual model in a D3.Js library, so that the visual effect of thermocline data is realized.
In order to verify the superiority of the ocean thermocline data visualization method based on the PCA-SVM, which is provided by the invention and is applied to an ocean thermocline, the models processed by the PCA and not processed by the PCA are compared under the condition of inputting the same thermocline data by combining with the current evaluation mechanism with authority comparison, in order to verify the superiority of the models compared with other algorithms, random forest algorithm, naive Bayes, KNN and logistic regression are respectively selected to test the same input ocean thermocline data, the comparison standards are operation time and accuracy, and the comparison result is shown in Table 1.
TABLE 1 comparison of the results of different algorithm model runs
Algorithm | Parameter(s) | Error of the first kind | Error of the second kind | Run time | Rate of accuracy |
SVM | λ=0.1,C=1 | 13 | 16 | 0.14960 | 99.55% |
BP-SVM | λ=0.1,C=1 | 2 | 9 | 0.09675 | 99.83% |
KNN | n_neighbors=5 | 9 | 5 | 0.19459 | 99.78% |
Naive Bayes | Auto | 25 | 16 | 0.01296 | 99.36% |
Logistic regression | Auto | 7 | 8 | 0.04199 | 99.77% |
Random forest | n_estimators=10 | 3 | 0 | 0.03191 | 99.95% |
Algorithm of the invention | λ=0.5,C=108.5 | 2 | 0 | 0.03291 | 99.97% |
As can be seen from Table 1, the models under different algorithms all obtain good operation indexes. However, by longitudinally comparing different models, it is obvious that the algorithm model provided by the invention has the optimal index, which shows that the ocean thermocline data visualization method based on the PCA-SVM provided by the invention has higher precision in the processing of the ocean thermocline data, and further, the data visualization also has higher display precision correspondingly.
In order to better implement the ocean thermocline data visualization method based on the PCA-SVM in the embodiment of the present invention, on the basis of the ocean thermocline data visualization method based on the PCA-SVM, correspondingly, the embodiment of the present invention further provides an ocean thermocline data visualization device based on the PCA-SVM, as shown in fig. 10, the ocean thermocline data visualization device 1000 based on the PCA-SVM includes:
a data acquisition module 1001 configured to acquire an initial ocean thermocline data set;
a dimensionality reduction processing module 1002, configured to perform dimensionality reduction processing on the initial ocean thermocline data based on a principal component analysis method, so as to obtain a dimensionality reduction ocean thermocline data set;
a filtering value processing module 1003, configured to perform filtering value processing on the dimensionality reduced ocean thermocline data based on a filtering control method, so as to obtain a filtered ocean thermocline data set;
the fitting processing module 1004 is used for acquiring a target BP neural network model, inputting the filtered ocean thermocline data set into the target BP neural network model for interpolation fitting to obtain a fitted ocean thermocline data set;
a classification processing module 1005, configured to obtain a target SVM model, input the fitted ocean thermocline data set into the target SVM model, and perform recognition and classification on ocean thermocline at different depths to obtain ocean thermocline target data;
and the visualization module 1006 is configured to perform visualization processing on the ocean thermocline target data to obtain a visualized ocean thermocline.
The ocean thermocline data visualization device 1000 based on the PCA-SVM provided in the above embodiment may implement the technical solutions described in the above ocean thermocline data visualization method based on the PCA-SVM, and the specific implementation principles of the above modules or units may refer to the corresponding contents in the above ocean thermocline data visualization method based on the PCA-SVM, which are not described herein again.
As shown in fig. 11, the present invention also provides an electronic device 1100 accordingly. The electronic device 1100 includes a processor 1101, a memory 1102, and a display 1103. Fig. 11 shows only some of the components of the electronic device 1100, but it is to be understood that not all of the shown components are required to be implemented, and that more or fewer components may be implemented instead.
The processor 1101 may be a Central Processing Unit (CPU), a microprocessor or other data Processing chip in some embodiments, and is used for executing program codes stored in the memory 1102 or Processing data, such as the ocean thermocline data visualization method based on PCA-SVM in the present invention.
In some embodiments, processor 1101 may be a single server or a group of servers. The server groups may be centralized or distributed. In some embodiments, the processor 1101 may be local or remote. In some embodiments, the processor 1101 may be implemented in a cloud platform. In an embodiment, the cloud platform may include a private cloud, a public cloud, a hybrid cloud, a community cloud, a distributed cloud, an intra-site, a multi-cloud, and the like, or any combination thereof.
The storage 1102 may in some embodiments be an internal storage unit of the electronic device 1100, such as a hard disk or a memory of the electronic device 1100. The memory 1102 may also be an external storage device of the electronic device 1100 in other embodiments, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), etc., provided on the electronic device 1100.
Further, the memory 1102 may also include both internal storage units and external storage devices of the electronic device 1100. The memory 1102 is used to store application software and various types of data for installing the electronic device 1100.
The display 1103 may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light-Emitting Diode) touch panel, or the like in some embodiments. The display 1103 is used to display information at the electronic device 1100 as well as to display a visual user interface. The components 1101-1103 of the electronic device 1100 communicate with each other via a system bus.
In an embodiment, when the processor 1101 executes the pipe network topology-based pipe leakage intelligent reorganization and recovery program in the memory 1102, the following steps may be implemented:
acquiring an initial ocean thermocline data set;
performing dimensionality reduction processing on the initial ocean thermocline data based on a principal component analysis method to obtain a dimensionality reduction ocean thermocline data set;
filtering the dimensionality-reduced ocean thermocline data based on a filtering control method to obtain a filtered ocean thermocline data set;
acquiring a target BP neural network model, inputting the filtered ocean thermocline data set into the target BP neural network model for interpolation fitting to obtain a fitted ocean thermocline data set;
acquiring a target SVM model, inputting the fitted ocean thermocline data set into the target SVM model to perform recognition classification of ocean thermocline at different depths to obtain ocean thermocline target data;
and carrying out visualization processing on the ocean thermocline target data to obtain a visualized ocean thermocline.
It should be understood that: the processor 1101, when executing the PCA-SVM based marine thermocline data visualization program in the memory 1102, may perform other functions in addition to the above functions, as described in the corresponding method embodiments above.
Further, the type of the mentioned electronic device 1100 is not specifically limited in the embodiments of the present invention, and the electronic device 1100 may be a portable electronic device such as a mobile phone, a tablet computer, a Personal Digital Assistant (PDA), a wearable device, and a laptop computer (laptop). Exemplary embodiments of portable electronic devices include, but are not limited to, portable electronic devices that carry an IOS, android, microsoft, or other operating system. The portable electronic device may also be other portable electronic devices such as laptop computers (laptop) with touch sensitive surfaces (e.g., touch panels) and the like. It should also be understood that in other embodiments of the present invention, the electronic device 1100 may not be a portable electronic device, but rather a desktop computer having a touch-sensitive surface (e.g., a touch panel).
Accordingly, the present application further provides a computer-readable storage medium, which is used for storing a computer-readable program or instruction, and when the program or instruction is executed by a processor, the step or the function in the PCA-SVM based ocean thermocline data visualization method provided by the above method embodiments can be implemented.
Those skilled in the art will appreciate that all or part of the flow of the method implementing the above embodiments may be implemented by instructing relevant hardware (such as a processor, a controller, etc.) by a computer program, and the computer program may be stored in a computer readable storage medium. The computer readable storage medium is a magnetic disk, an optical disk, a read-only memory or a random access memory.
The method and the device for visualizing the ocean thermocline data based on the PCA-SVM provided by the invention are described in detail, specific examples are applied in the method to explain the principle and the implementation mode of the invention, and the description of the examples is only used for helping to understand the method and the core idea of the invention; meanwhile, for those skilled in the art, according to the idea of the present invention, there may be variations in the specific embodiments and the application scope, and in summary, the content of the present specification should not be construed as a limitation to the present invention.
Claims (10)
1. A visualization method for ocean thermocline data based on PCA-SVM is characterized by comprising the following steps:
acquiring an initial ocean thermocline data set;
performing dimensionality reduction processing on the initial ocean thermocline data based on a principal component analysis method to obtain a dimensionality reduction ocean thermocline data set;
filtering the dimensionality-reduced ocean thermocline data based on a filtering control method to obtain a filtered ocean thermocline data set;
acquiring a target BP neural network model, inputting the filtered ocean thermocline data set into the target BP neural network model for interpolation fitting to obtain a fitted ocean thermocline data set;
acquiring a target SVM model, inputting the fitted ocean thermocline data set into the target SVM model to perform recognition classification of ocean thermocline at different depths to obtain ocean thermocline target data;
and carrying out visualization processing on the target data of the ocean thermocline to obtain a visualized ocean thermocline.
2. The PCA-SVM based marine thermocline data visualization method according to claim 1, wherein said initial marine thermocline data set comprises four dimensions and said four dimensions correspond to sea temperature data, wherein said four dimensions comprise longitude, latitude, depth and time.
3. The method for visualizing ocean thermocline data based on PCA-SVM of claim 1, wherein the dimensionality reduction of the initial ocean thermocline data set based on principal component analysis to obtain a dimensionality reduced ocean thermocline data set comprises:
carrying out zero-valued processing on the initial ocean thermocline data set to obtain a centralized matrix;
calculating covariance according to the centralized matrix to obtain a covariance matrix;
obtaining a covariance coefficient calculation eigenvalue according to the covariance matrix, and calculating an eigenvector according to the eigenvalue to construct an eigenvector matrix;
screening the characteristic vector matrix according to a preset characteristic value screening threshold value to obtain a primary screening matrix;
and carrying out scaling processing on the primary screening matrix to obtain a scaling matrix, wherein the scaling matrix forms the dimensionality reduction ocean thermocline data set.
4. The ocean thermocline data visualization method based on PCA-SVM according to claim 1, wherein the filtering control method based on filtering value processing is performed on the reduced-dimension ocean thermocline data to obtain a filtered ocean thermocline data set, comprising:
carrying out abnormal data filtering value processing on the dimensionality-reduced ocean thermocline data set to obtain a preliminary filtering value ocean thermocline data set;
and performing mean square error calculation on the data which is not subjected to the filtering value processing in the preliminary filtering value ocean thermocline data set, and performing exception marking on the data of which the difference between the data value and the average value in the data which is not subjected to the filtering value processing is larger than a preset multiplying power mean square error to obtain the filtering ocean thermocline data set.
5. The method of visualizing ocean thermocline data based on PCA-SVM of claim 1, wherein determining the target BP neural network model comprises:
acquiring a data sample set of an ocean thermocline to be fitted;
determining the number of input layers, hidden layers and output layers in the BP neural network, and establishing an initial BP neural network model;
determining initial weight values and initial threshold values of the input layer, the hidden layer and the output layer, and performing iterative training on the initial BP neural network model based on the ocean thermocline data sample set to be fitted;
and updating the initial weight values and the initial threshold values of the input layer, the hidden layer and the output layer based on a genetic algorithm until the preset training precision is reached to obtain the target BP neural network model.
6. The PCA-SVM based marine thermocline data visualization method according to claim 1, wherein determining the target SVM model comprises:
obtaining an ocean thermocline data sample set to be classified, and selecting different proportions to divide the ocean thermocline data sample set to be classified into a training set and a testing set;
determining an optimization objective function, a model training function, a decision function and an SVM kernel function to establish an initial SVM model;
and respectively carrying out training test on the SVM model based on the training set and the test set until an optimization punishment factor and a kernel function coefficient are determined to obtain a target SVM model.
7. The ocean thermocline data visualization method based on PCA-SVM according to claim 1, wherein the visualizing the target ocean thermocline data set and importing a visualization model to realize the visualization of the thermocline data comprises the following steps:
performing matrix transformation on the target ocean thermocline data set to obtain a one-dimensional json data format file;
adding header file information to the one-dimensional json data format file, wherein the header file information comprises a time longitude and latitude range, time, ocean depth and a display mode to obtain a target json file;
and importing the target json file into a visualization model to realize the visualization of the ocean thermocline data.
8. An ocean thermocline data visualization device based on PCA-SVM is characterized by comprising the following components:
the data acquisition module is used for acquiring an initial ocean thermocline data set;
the dimensionality reduction processing module is used for carrying out dimensionality reduction processing on the initial ocean thermocline data based on a principal component analysis method to obtain a dimensionality reduction ocean thermocline data set;
the filtering value processing module is used for carrying out filtering value processing on the dimensionality reduction ocean thermocline data based on a filtering control method to obtain a filtering ocean thermocline data set;
the fitting processing module is used for acquiring a target BP neural network model, inputting the filtered ocean thermocline data set into the target BP neural network model for interpolation fitting to obtain a fitted ocean thermocline data set;
the classification processing module is used for acquiring a target SVM model, inputting the fitted ocean thermocline data set into the target SVM model to perform recognition and classification of ocean thermocline at different depths to obtain ocean thermocline target data;
and the visualization module is used for carrying out visualization processing on the ocean thermocline target data to obtain a visualized ocean thermocline.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor, when executing the program, implements the PCA-SVM based marine thermocline data visualization method according to any one of claims 1 to 7.
10. A computer-readable storage medium, on which a computer program is stored, which program, when being executed by a processor, carries out a method for visualizing ocean thermocline data based on PCA-SVM according to any one of claims 1 to 7.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202211022965.1A CN115358327A (en) | 2022-08-24 | 2022-08-24 | Ocean thermocline data visualization method, device, equipment and medium based on PCA-SVM |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202211022965.1A CN115358327A (en) | 2022-08-24 | 2022-08-24 | Ocean thermocline data visualization method, device, equipment and medium based on PCA-SVM |
Publications (1)
Publication Number | Publication Date |
---|---|
CN115358327A true CN115358327A (en) | 2022-11-18 |
Family
ID=84004012
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202211022965.1A Pending CN115358327A (en) | 2022-08-24 | 2022-08-24 | Ocean thermocline data visualization method, device, equipment and medium based on PCA-SVM |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN115358327A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116028510A (en) * | 2023-03-30 | 2023-04-28 | 国家海洋技术中心 | Ocean jump layer searching method and system |
-
2022
- 2022-08-24 CN CN202211022965.1A patent/CN115358327A/en active Pending
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116028510A (en) * | 2023-03-30 | 2023-04-28 | 国家海洋技术中心 | Ocean jump layer searching method and system |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Shi et al. | Development of subsurface geological cross-section from limited site-specific boreholes and prior geological knowledge using iterative convolution XGBoost | |
Fayyad et al. | KDD for Science Data Analysis: Issues and Examples. | |
Filzmoser et al. | Identification of local multivariate outliers | |
Willett et al. | Minimax optimal level-set estimation | |
Chen et al. | A feature discretization method for classification of high-resolution remote sensing images in coastal areas | |
CN116665067B (en) | Ore finding target area optimization system and method based on graph neural network | |
Foster et al. | Probabilistic machine learning estimation of ocean mixed layer depth from dense satellite and sparse in situ observations | |
CN115358327A (en) | Ocean thermocline data visualization method, device, equipment and medium based on PCA-SVM | |
CN114972737A (en) | Remote sensing image target detection system and method based on prototype comparison learning | |
Zhang et al. | SaltISCG: Interactive salt segmentation method based on CNN and graph cut | |
CN114722926A (en) | Convolution clustering method for scale vortex time sequence diagram in towed sensor array | |
Huang et al. | A multivariable optical remote sensing image feature discretization method applied to marine vessel targets recognition | |
CN116927771A (en) | Method, device, equipment and medium for predicting total organic carbon data of shale reservoir | |
Dubey et al. | Usage of clustering and weighted nearest neighbors for efficient missing data imputation of microarray gene expression dataset | |
CN114402233A (en) | Automatic calibration of forward deposition model | |
CN114860953B (en) | Interpretable method based on less-sample relation prediction model | |
CN116337377A (en) | Multi-sensor-based mining equipment fault state analysis method and device | |
CN112541292B (en) | Submarine cable burial depth estimation algorithm based on distributed optical fiber temperature measurement principle | |
Liu et al. | Mining weather data using fuzzy cluster analysis | |
Buaba et al. | Locality sensitive hashing for satellite images using texture feature vectors | |
CN115631799B (en) | Sample phenotype prediction method and device, electronic equipment and storage medium | |
CN111542819A (en) | Apparatus and method for improved subsurface data processing system | |
CN117421642B (en) | Deep learning-based intelligent detector data storage method and related equipment | |
CN115879034B (en) | Tropical cyclone strength monitoring method, device and equipment based on machine learning | |
CN114580372B (en) | Text processing method, device, equipment and storage medium |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination |