CN112990789B - User health risk analysis system - Google Patents

User health risk analysis system Download PDF

Info

Publication number
CN112990789B
CN112990789B CN202110503792.4A CN202110503792A CN112990789B CN 112990789 B CN112990789 B CN 112990789B CN 202110503792 A CN202110503792 A CN 202110503792A CN 112990789 B CN112990789 B CN 112990789B
Authority
CN
China
Prior art keywords
data
elements
matrix
absolute value
data elements
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.)
Active
Application number
CN202110503792.4A
Other languages
Chinese (zh)
Other versions
CN112990789A (en
Inventor
姚娟娟
樊代明
钟南山
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shanghai Mingping Medical Data Technology Co ltd
Original Assignee
Mingpinyun Beijing Data Technology Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Mingpinyun Beijing Data Technology Co Ltd filed Critical Mingpinyun Beijing Data Technology Co Ltd
Priority to CN202110503792.4A priority Critical patent/CN112990789B/en
Publication of CN112990789A publication Critical patent/CN112990789A/en
Application granted granted Critical
Publication of CN112990789B publication Critical patent/CN112990789B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0635—Risk analysis of enterprise or organisation activities
    • 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
    • G06N3/045—Combinations of networks
    • 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
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
    • G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Human Resources & Organizations (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Strategic Management (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Economics (AREA)
  • Educational Administration (AREA)
  • Development Economics (AREA)
  • Software Systems (AREA)
  • Artificial Intelligence (AREA)
  • General Engineering & Computer Science (AREA)
  • Molecular Biology (AREA)
  • Mathematical Physics (AREA)
  • General Health & Medical Sciences (AREA)
  • Evolutionary Computation (AREA)
  • Data Mining & Analysis (AREA)
  • Computational Linguistics (AREA)
  • Biophysics (AREA)
  • Biomedical Technology (AREA)
  • Computing Systems (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Game Theory and Decision Science (AREA)
  • Health & Medical Sciences (AREA)
  • Marketing (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Tourism & Hospitality (AREA)
  • General Business, Economics & Management (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

The invention provides a user health risk analysis system, and a user health risk analysis method comprises the following steps: acquiring user data, wherein the user data comprises a plurality of data elements, and comparing each data element with corresponding reference data to determine multidimensional data elements; determining the positions of matrix elements corresponding to a plurality of data elements in a feature matrix, and respectively loading the multidimensional data elements into the matrix elements of the feature matrix; inputting the characteristic matrix corresponding to the multidimensional data elements into a neural network unit for training to obtain an analysis model; and inputting the feature matrix to be analyzed into the analysis model, acquiring a risk feature vector, and determining a user risk expected value through the risk feature vector. And obtaining the incidence relation and the quantitative index between the multidimensional data elements and the user risk by performing coupling analysis on the multidimensional data elements.

Description

User health risk analysis system
Technical Field
The invention relates to the technical field of artificial intelligence, in particular to a user health risk analysis system.
Background
In the daily life, work, rest or physical examination process, a large amount of user data is generated by a user, various risk characteristic information such as the health and potential risks of the user is often hidden in the user data, and different types of user data have complex relevance, multi-coupling and non-dominance, so that a mining method for acquiring deep risk characteristic information by processing the represented user data and a method for evaluating and controlling the risks of the user through the risk characteristic information are lacked.
Disclosure of Invention
In view of the above-mentioned shortcomings of the prior art, the present invention provides a method, a system, an electronic device and a medium for analyzing user health risks, which are used to solve the problem of analyzing user data of an image in the prior art.
To achieve the above and other related objects, the present invention provides a method for analyzing health risks of a user, including:
acquiring user data, wherein the user data comprises a plurality of data elements, comparing each data element with corresponding reference data to determine multidimensional data elements, and the multidimensional data elements at least comprise one of the following data elements: standard data elements, standard exceeding data elements and non-standard data elements;
determining the positions of matrix elements corresponding to a plurality of data elements in a feature matrix, and respectively loading the multidimensional data elements into the matrix elements of the feature matrix;
inputting the characteristic matrix corresponding to the multidimensional data elements into a neural network unit for training to obtain an analysis model;
and inputting the feature matrix to be analyzed into the analysis model, acquiring a risk feature vector, and determining a user risk expected value through the risk feature vector.
Optionally, the step of loading the multidimensional data elements into matrix elements of the feature matrix respectively includes:
loading the qualified data elements into matrix elements of the feature matrix;
determining a reference interpolation according to the reference data, and acquiring a standard-reaching parameter through a ratio of a first absolute value and a second absolute value, wherein the first absolute value is an absolute value of a difference value between the standard-reaching data element and the reference interpolation, and the second absolute value is an absolute value of a difference value between two endpoint values in the reference data;
and loading the standard-reaching parameters into matrix elements of the feature matrix to obtain a first feature matrix.
Optionally, the reference interpolation is a median or an average between two end point values in the reference data.
Optionally, the step of loading the multidimensional data elements into matrix elements of the feature matrix respectively includes:
loading the superscalar data elements into matrix elements of the feature matrix;
acquiring a standard exceeding index through a ratio of a third absolute value and a second absolute value, and determining a standard exceeding parameter through the standard exceeding index and a logarithm, wherein the second absolute value is an absolute value of a difference value of two endpoint values in the reference data, and the third absolute value is an absolute value of a difference value of the standard exceeding data element and one endpoint value of the reference data;
and loading the standard exceeding parameters into matrix elements of the feature matrix to obtain a second feature matrix.
Optionally, the step of loading the multidimensional data elements into matrix elements of the feature matrix respectively includes:
loading the substandard data elements into matrix elements of the feature matrix;
obtaining an unqualified index according to the ratio of a fourth absolute value and a second absolute value, and determining an unqualified parameter according to the unqualified index and a logarithm, wherein the second absolute value is the absolute value of the difference between two end points in the reference data, and the fourth absolute value is the absolute value of the difference between the unqualified data element and an end point of the reference data;
and loading the substandard parameters into matrix elements of the feature matrix to obtain a third feature matrix.
Optionally, the step of inputting the feature matrix corresponding to the multidimensional data element into a neural network unit for training, and obtaining an analysis model includes:
the neural network unit comprises a first neural network, a second neural network and a third neural network, the characteristic matrix corresponding to the standard data elements is input into the first neural network, the characteristic matrix corresponding to the standard data elements is input into the second neural network, and the characteristic matrix corresponding to the substandard data elements is input into the third neural network;
splicing the output values of the first neural network, the second neural network and the third neural network to obtain the output value of the neural network unit, wherein the output value of the neural network unit is a user risk prediction value;
and training the neural network unit through an average absolute error loss function, improving the confidence coefficient of the user risk prediction value, and acquiring an analysis model when the confidence coefficient reaches a preset value.
Optionally, the first neural network, the second neural network, and the third neural network each include: an input layer, a hidden layer, and an output layer.
A user health risk analysis system, comprising:
the acquisition module is used for acquiring user data, the user data comprises a plurality of data elements, each data element is compared with corresponding reference data, and a multidimensional data element is determined, wherein the multidimensional data element at least comprises one of the following data elements: standard data elements, standard exceeding data elements and non-standard data elements;
the preprocessing module is used for determining the positions of matrix elements corresponding to a plurality of data elements in a characteristic matrix and respectively loading the multidimensional data elements into the matrix elements of the characteristic matrix;
the model module is used for inputting the feature matrix corresponding to the multidimensional data elements into a neural network unit for training to obtain an analysis model;
the analysis module is used for inputting the feature matrix to be analyzed into the analysis model, acquiring a risk feature vector and determining a user risk expected value through the risk feature vector;
the acquisition module, the preprocessing module, the model module and the analysis module are in signal connection.
An electronic device, comprising: one or more processors; and one or more machine readable media having instructions stored thereon that, when executed by the one or more processors, cause the electronic device to perform one or more of the methods.
A machine-readable medium having stored thereon instructions, which when executed by one or more processors, cause an apparatus to perform one or more of the methods described.
As described above, the user health risk analysis method, system, electronic device, and medium of the present invention have the following beneficial effects:
and obtaining the incidence relation and the quantitative index between the multidimensional data elements and the user risk by performing coupling analysis on the multidimensional data elements.
Drawings
Fig. 1 is a schematic flow chart illustrating a user health risk analysis method according to an embodiment of the present invention.
Fig. 2 is a schematic structural diagram of a feature matrix according to an embodiment of the invention.
FIG. 3 is a flowchart illustrating feature matrix processing according to an embodiment of the invention.
Fig. 4 is a schematic structural diagram of a user health risk analysis system according to an embodiment of the present invention.
Detailed Description
The embodiments of the present invention are described below with reference to specific embodiments, and other advantages and effects of the present invention will be easily understood by those skilled in the art from the disclosure of the present specification. The invention is capable of other and different embodiments and of being practiced or of being carried out in various ways, and its several details are capable of modification in various respects, all without departing from the spirit and scope of the present invention.
It should be noted that the drawings provided in the present embodiment are only for illustrating the basic idea of the present invention, and the components related to the present invention are only shown in the drawings rather than drawn according to the number, shape and size of the components in actual implementation, and the type, quantity and proportion of the components in actual implementation may be changed freely, and the layout of the components may be more complicated. The structures, proportions, sizes, and other dimensions shown in the drawings and described in the specification are for understanding and reading the present disclosure, and are not intended to limit the scope of the present disclosure, which is defined in the claims, and are not essential to the art, and any structural modifications, changes in proportions, or adjustments in size, which do not affect the efficacy and attainment of the same are intended to fall within the scope of the present disclosure. In addition, the terms "upper", "lower", "left", "right", "middle" and "one" used in the present specification are for clarity of description, and are not intended to limit the scope of the present invention, and the relative relationship between the terms and the terms is not to be construed as a scope of the present invention.
The inventor finds that a large amount of user data can be generated in daily life, work, physical examination or physical examination of a user, and the ways of acquiring the user data are increasing, for example, the data can be acquired from a body fat scale and wearable equipment, or from historical physical examination or physical examination data of the user, data elements in the user data generally correspond to reference data so that the state of quantitative measurement data elements is up to standard, over-standard or not up to standard, and the incidence relation between the multidimensional data elements and the user risk and the quantitative indexes are acquired by performing coupling analysis on the multidimensional data elements And mining and analyzing to obtain a user risk predicted value with higher confidence. Referring to fig. 1, the present invention provides a method for analyzing a health risk of a user, including:
s1: acquiring user data, wherein the user data comprises a plurality of data elements, the user data corresponding to each user ID comprises a plurality of types of data elements, the plurality of types of data elements are related to the user information, comparing each data element with corresponding reference data, and determining a multi-dimensional data element, and the multi-dimensional data element at least comprises one of the following data elements: the inventor also finds that the influence of the over-standard data elements and the under-standard data elements on the user risk is relatively large, and the influence of the up-standard data elements on the user risk is relatively small;
s2: determining the positions of matrix elements corresponding to a plurality of data elements in a feature matrix, respectively loading the multidimensional data elements into the matrix elements of the feature matrix, respectively matrixing standard data elements, standard exceeding data elements and substandard data elements, and bearing deep feature information of user data through the matrixed multidimensional data elements;
s3: inputting the feature matrix corresponding to the multidimensional data elements into a neural network unit for training to obtain an analysis model, wherein due to different influences of the standard data elements, standard exceeding data elements and substandard data elements on user risks, the multidimensional data is input for coupling analysis to improve the confidence coefficient of risk prediction;
s4: and inputting the feature matrix to be analyzed into the analysis model, acquiring a risk feature vector, and determining a user risk expected value through the risk feature vector.
The inventor finds that values of some reference data are wide, although data elements are in a standard reaching state, the reference data are possibly close to end points of the reference data, when a plurality of data elements are close to the end points of the reference data, risk probability of a user is relatively high, but influence indexes of user risks cannot be obtained from marking data elements of the representation, in order to obtain incidence relation and influence weight of the standard reaching data to the user risks from the standard reaching data elements, the inventor provides a method for processing the standard reaching data elements, and in some implementation processes, the step of respectively loading the multidimensional data elements into matrix elements of the feature matrix comprises the following steps:
loading the qualified data elements D into matrix elements of the feature matrix, the structure of the feature matrix referring to fig. 2, for example, a11, … …, A1n, … …, and Amn may be qualified data elements, and may be superstandard data elements or substandard data elements, the size of the feature matrix may depend on the capacity of the data elements, when the capacity of the data elements is insufficient, the unloaded matrix elements may be filled with 0, and the matrix elements of each feature matrix may be loaded with a type of data elements, such as qualified data elements, superstandard data elements, or substandard data elements;
determining a reference interpolation A according to the reference data, and obtaining an up-to-standard parameter D = B/C through a ratio B/C of a first absolute value B and a second absolute value C, wherein the first absolute value B is an absolute value of a difference value between the up-to-standard data element a and the reference interpolation A, the second absolute value C is an absolute value of a difference value between two endpoint values E and F in the reference data, and C = | E-F |;
and loading the standard-reaching parameter D into the matrix elements of the characteristic matrix to obtain a first characteristic matrix, and loading the other matrix elements into 0.
In order to reflect the characteristic information of the standard data element more intuitively, a median or an average between two end point values in the reference data can be selected as the reference interpolation, for example, when the reference data is a discrete data set, the median can be selected as the reference interpolation, and for example, when the reference data is a continuous data set, the average can be selected as the reference interpolation.
The inventor finds that the influence of the exceeding degree or the exceeding proportion in the exceeding data on the user risk is not a linear change process, so that the inventor provides a method for processing the exceeding data elements, and the influence of the exceeding degree or the exceeding proportion of the exceeding data elements on the user risk can be reflected in the processed exceeding data elements. Illustratively, the step of loading the multidimensional data elements into matrix elements of the feature matrix, respectively, comprises:
loading the superscalar data elements L into matrix elements of the feature matrix;
obtaining the over-standard index G = H/C through the ratio H/C of the third absolute value H and the second absolute value C, and determining the over-standard parameter e through the over-standard index G and the logarithm which can select a natural logarithm e (e ≈ 2.7182818284)GAnd selecting a proper logarithm according to the importance of the superscalar data element, wherein the logarithm can also be a logarithm with the base of pi, wherein the second absolute value C is the absolute value of the difference value of two endpoint values E and F in the reference data, C = | E-F |,the third absolute value H is an absolute value of a difference E or F between the superscalar data element L and an endpoint value of the reference data, where H = | L-E | or H = | L-F |;
the superscalar parameter e may be scaledGAnd loading the matrix elements into the characteristic matrix to obtain a second characteristic matrix, and loading the rest matrix elements into 0. The dominant result of the exceeding data elements on the user risk influence is obtained by carrying out exponential operation on the exceeding proportion of the exceeding data elements, the purpose of increasing the weight of the exceeding data elements on the user risk influence is achieved, and the accuracy and the confidence coefficient of the user risk prediction are improved.
The inventor also finds that the influence of the non-standard degree or the non-standard proportion in the non-standard data on the user risk is not a linear change process, and the non-standard data has strong correlation with the user risk state, so that the inventor provides a method for processing the non-standard data elements, and the influence of the non-standard degree or the non-standard proportion of the non-standard data elements on the user risk can be reflected in the processed non-standard data elements. The step of loading the multidimensional data elements into the matrix elements of the feature matrix, respectively, includes:
loading the substandard data elements Z into matrix elements of the feature matrix;
obtaining the unqualified index Y = X/C through the ratio X/C of the fourth absolute value X and the second absolute value C, and determining the standard exceeding parameter e through the unqualified index Y and the logarithm which can select a natural logarithm e (e ≈ 2.7182818284)GSelecting a suitable logarithm depending on the importance of the substandard data element, and selecting a logarithm with pi as a base, wherein the second absolute value C is an absolute value of a difference between two end-point values E and F in the reference data, the fourth absolute value X is an absolute value of a difference between the substandard data element Z and an end-point value E or F of the reference data, and X = | Z-E | or H = | Z-F |;
the non-compliance parameter e may be setYLoading the feature matrix into matrix elements of the feature matrix to obtain a third feature matrix, and loading the rest matricesThe element is loaded with 0.
In order to further perform coupling analysis on multidimensional data elements, a feature matrix corresponding to the multidimensional data elements is trained by a machine learning method, and is learned therefrom to obtain a risk feature vector related to a user risk, please refer to fig. 4, the feature matrix corresponding to the multidimensional data elements is input into a neural network unit for training, and the step of obtaining an analysis model includes:
the neural network unit comprises a first neural network 10, a second neural network 20 and a third neural network 30, the characteristic matrix corresponding to the standard data elements is input into the first neural network 10, the characteristic matrix corresponding to the standard data elements is input into the second neural network 20, and the characteristic matrix corresponding to the substandard data elements is input into the third neural network 30;
the output values of the first neural network 10, the second neural network 20 and the third neural network 30 are spliced to obtain the output values of the neural network units, the output values of the neural network units are predicted user risk values, and the spliced output values of the first neural network 10, the second neural network 20 and the third neural network 30 can be further input into a fourth neural network 40 before the output values of the neural network units are obtained;
training the neural network unit through an average Absolute Error Loss function, improving the confidence coefficient of the user risk prediction value, and acquiring an analysis model when the confidence coefficient reaches a preset value, wherein the average Absolute Error Loss function (Mean Absolute Error/MAE) is a common Loss function, and is also called as an L1 Loss function.
To facilitate training and coupling analysis of the feature matrix, the first, second, and third neural networks each include: an input layer 10, a hidden layer 20 and an output layer 30. Further, the fourth neural network 40 also includes an input layer, a hidden layer, and an output layer. The hidden layer comprises a plurality of layers and each time comprises a plurality of neuron nodes, the hidden layers of the first neural network, the second neural network, the third neural network and the fourth neural network are trained through an average absolute value error loss function, the weight value input by the neuron nodes of each hidden layer is adjusted through repeated iterative training of forward propagation and backward propagation, an optimized user risk predicted value is obtained, and when the user risk predicted value reaches high confidence coefficient, the user risk predicted value can measure the general situation of user risk.
The mathematical expression of the user risk expectation value determined by the risk feature vector is as follows:
Figure 742535DEST_PATH_IMAGE001
wherein the risk feature vector is an i-dimensional vector (t)i) I is not less than 1 and i is a positive integer, tiIs the probability value of the user risk in the ith year, xiAnd Q is a gain coefficient, and M is a user risk expected value. The user can obtain the predicted risk condition in a future period of time through the risk feature vector, for example, the expected risk value of the user in 1 st year, 2 nd year, 3 rd year, … … th year and i th year is fed back through the i-dimensional vector, so that the expected risk condition of the user can be reasonably evaluated, further health detection or physical detection can be made, or corresponding intervention decision can be made.
Referring to fig. 4, the present invention further provides a system for analyzing health risks of users, including:
the acquisition module is used for acquiring user data, the user data comprises a plurality of data elements, each data element is compared with corresponding reference data, and a multidimensional data element is determined, wherein the multidimensional data element at least comprises one of the following data elements: standard data elements, standard exceeding data elements and non-standard data elements;
the preprocessing module is used for determining the positions of matrix elements corresponding to a plurality of data elements in a characteristic matrix and respectively loading the multidimensional data elements into the matrix elements of the characteristic matrix;
the model module is used for inputting the feature matrix corresponding to the multidimensional data elements into a neural network unit for training to obtain an analysis model;
the analysis module is used for inputting the feature matrix to be analyzed into the analysis model, acquiring a risk feature vector and determining a user risk expected value through the risk feature vector;
the acquisition module, the preprocessing module, the model module and the analysis module are in signal connection. And obtaining the incidence relation and the quantitative index between the multidimensional data elements and the user risk by performing coupling analysis on the multidimensional data elements.
Optionally, the step of loading the multidimensional data elements into matrix elements of the feature matrix respectively includes:
loading the qualified data elements into matrix elements of the feature matrix;
determining a reference interpolation according to the reference data, and acquiring a standard-reaching parameter through a ratio of a first absolute value and a second absolute value, wherein the first absolute value is an absolute value of a difference value between the standard-reaching data element and the reference interpolation, and the second absolute value is an absolute value of a difference value between two endpoint values in the reference data;
and loading the standard-reaching parameters into matrix elements of the feature matrix to obtain a first feature matrix.
Optionally, the reference interpolation is a median or an average between two end point values in the reference data.
Optionally, the step of loading the multidimensional data elements into matrix elements of the feature matrix respectively includes:
loading the superscalar data elements into matrix elements of the feature matrix;
acquiring a standard exceeding index through a ratio of a third absolute value and a second absolute value, and determining a standard exceeding parameter through the standard exceeding index and a logarithm, wherein the second absolute value is an absolute value of a difference value of two endpoint values in the reference data, and the third absolute value is an absolute value of a difference value of the standard exceeding data element and one endpoint value of the reference data;
and loading the standard exceeding parameters into matrix elements of the feature matrix to obtain a second feature matrix.
Optionally, the step of loading the multidimensional data elements into matrix elements of the feature matrix respectively includes:
loading the substandard data elements into matrix elements of the feature matrix;
obtaining an unqualified index according to the ratio of a fourth absolute value and a second absolute value, and determining an unqualified parameter according to the unqualified index and a logarithm, wherein the second absolute value is the absolute value of the difference between two end points in the reference data, and the fourth absolute value is the absolute value of the difference between the unqualified data element and an end point of the reference data;
and loading the substandard parameters into matrix elements of the feature matrix to obtain a third feature matrix.
Optionally, the step of inputting the feature matrix corresponding to the multidimensional data element into a neural network unit for training, and obtaining an analysis model includes:
the neural network unit comprises a first neural network, a second neural network and a third neural network, the characteristic matrix corresponding to the standard data elements is input into the first neural network, the characteristic matrix corresponding to the standard data elements is input into the second neural network, and the characteristic matrix corresponding to the substandard data elements is input into the third neural network;
splicing the output values of the first neural network, the second neural network and the third neural network to obtain the output value of the neural network unit, wherein the output value of the neural network unit is a user risk prediction value;
and training the neural network unit through an average absolute error loss function, improving the confidence coefficient of the user risk prediction value, and acquiring an analysis model when the confidence coefficient reaches a preset value.
Optionally, the first neural network, the second neural network, and the third neural network each include: an input layer, a hidden layer, and an output layer.
An embodiment of the present invention provides an electronic device, including: one or more processors; and one or more machine readable media having instructions stored thereon that, when executed by the one or more processors, cause the electronic device to perform one or more of the methods. The invention is operational with numerous general purpose or special purpose computing system environments or configurations. For example: personal computers, server computers, hand-held or portable devices, tablet-type devices, multiprocessor systems, microprocessor-based systems, set top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, distributed computing environments that include any of the above systems or devices, and the like.
Embodiments of the invention also provide one or more machine-readable media having instructions stored thereon, which when executed by one or more processors, cause an apparatus to perform one or more of the methods described herein. The invention may be described in the general context of computer-executable instructions, such as program modules, being executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular abstract data types. The invention may also be practiced in distributed computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote computer storage media including memory storage devices.
The foregoing embodiments are merely illustrative of the principles and utilities of the present invention and are not intended to limit the invention. Any person skilled in the art can modify or change the above-mentioned embodiments without departing from the spirit and scope of the present invention. Accordingly, it is intended that all equivalent modifications or changes which can be made by those skilled in the art without departing from the spirit and technical spirit of the present invention be covered by the claims of the present invention.

Claims (1)

1. A user health risk analysis system, comprising:
the acquisition module is used for acquiring user data, the user data comprises a plurality of data elements, each data element is compared with corresponding reference data, and a multidimensional data element is determined, wherein the multidimensional data element at least comprises one of the following data elements: the system comprises standard data elements, standard exceeding data elements and non-standard data elements, wherein user data are obtained from body fat scales, wearable equipment, and historical physical examination or physical measurement data of a user;
a preprocessing module, configured to determine positions of matrix elements corresponding to a plurality of data elements in a feature matrix, load the multidimensional data elements into the matrix elements of the feature matrix, and load the multidimensional data elements into the matrix elements of the feature matrix, respectively, including: loading the qualified data elements into matrix elements of the feature matrix; determining a reference interpolation according to the reference data, and acquiring a standard-reaching parameter through a ratio of a first absolute value and a second absolute value, wherein the first absolute value is an absolute value of a difference value between the standard-reaching data element and the reference interpolation, and the second absolute value is an absolute value of a difference value between two endpoint values in the reference data; loading the standard-reaching parameters into matrix elements of the feature matrix to obtain a first feature matrix; loading the superscalar data elements into matrix elements of the feature matrix; acquiring a standard exceeding index through a ratio of a third absolute value and a second absolute value, and determining a standard exceeding parameter through the standard exceeding index and a logarithm, wherein the second absolute value is an absolute value of a difference value of two endpoint values in the reference data, and the third absolute value is an absolute value of a difference value of the standard exceeding data element and one endpoint value of the reference data; loading the standard exceeding parameters into matrix elements of the feature matrix to obtain a second feature matrix; loading the substandard data elements into matrix elements of the feature matrix; obtaining an unqualified index according to the ratio of a fourth absolute value and a second absolute value, and determining an unqualified parameter according to the unqualified index and a logarithm, wherein the second absolute value is the absolute value of the difference between two end points in the reference data, and the fourth absolute value is the absolute value of the difference between the unqualified data element and an end point of the reference data; loading the substandard parameters into matrix elements of the feature matrix to obtain a third feature matrix;
the model module is used for inputting the feature matrix corresponding to the multidimensional data elements into a neural network unit for training to obtain an analysis model;
the analysis module is used for inputting the feature matrix to be analyzed into the analysis model, acquiring a risk feature vector and determining a user risk expected value through the risk feature vector;
the acquisition module, the preprocessing module, the model module and the analysis module are in signal connection.
CN202110503792.4A 2021-05-10 2021-05-10 User health risk analysis system Active CN112990789B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110503792.4A CN112990789B (en) 2021-05-10 2021-05-10 User health risk analysis system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110503792.4A CN112990789B (en) 2021-05-10 2021-05-10 User health risk analysis system

Publications (2)

Publication Number Publication Date
CN112990789A CN112990789A (en) 2021-06-18
CN112990789B true CN112990789B (en) 2021-11-02

Family

ID=76337412

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110503792.4A Active CN112990789B (en) 2021-05-10 2021-05-10 User health risk analysis system

Country Status (1)

Country Link
CN (1) CN112990789B (en)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1738426A (en) * 2005-09-09 2006-02-22 南京大学 Video motion goal division and track method
CN103544397A (en) * 2013-10-30 2014-01-29 江苏万全科技有限公司 Summarizing, analyzing and warning method for human physical health index data
CN110720918A (en) * 2012-05-30 2020-01-24 瑞思迈传感器技术有限公司 Method and apparatus for monitoring cardiopulmonary health
CN111951954A (en) * 2020-08-10 2020-11-17 中国平安人寿保险股份有限公司 Body health state detection method and device, readable storage medium and terminal equipment
CN112002429A (en) * 2020-10-27 2020-11-27 北京梦天门科技股份有限公司 Physical examination portrait and service recommendation system and method for public health enterprise personnel
CN112687374A (en) * 2021-01-12 2021-04-20 湖南师范大学 Psychological crisis early warning method based on text and image information joint calculation

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108249544B (en) * 2018-02-07 2023-12-19 华中科技大学 Arsenic-containing wastewater treatment method and device
CN111652627B (en) * 2020-07-07 2024-04-23 中国银行股份有限公司 Risk assessment method and risk assessment device

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1738426A (en) * 2005-09-09 2006-02-22 南京大学 Video motion goal division and track method
CN110720918A (en) * 2012-05-30 2020-01-24 瑞思迈传感器技术有限公司 Method and apparatus for monitoring cardiopulmonary health
CN103544397A (en) * 2013-10-30 2014-01-29 江苏万全科技有限公司 Summarizing, analyzing and warning method for human physical health index data
CN111951954A (en) * 2020-08-10 2020-11-17 中国平安人寿保险股份有限公司 Body health state detection method and device, readable storage medium and terminal equipment
CN112002429A (en) * 2020-10-27 2020-11-27 北京梦天门科技股份有限公司 Physical examination portrait and service recommendation system and method for public health enterprise personnel
CN112687374A (en) * 2021-01-12 2021-04-20 湖南师范大学 Psychological crisis early warning method based on text and image information joint calculation

Also Published As

Publication number Publication date
CN112990789A (en) 2021-06-18

Similar Documents

Publication Publication Date Title
CN112598111B (en) Abnormal data identification method and device
Bonassi et al. Bayesian learning from marginal data in bionetwork models
CN114330713A (en) Convolutional neural network model pruning method and device, electronic equipment and storage medium
CN117349798B (en) Unbalanced regression-based mechanical drilling rate prediction method and system
CN113743971A (en) Data processing method and device
CN112200296A (en) Network model quantification method and device, storage medium and electronic equipment
CN117390635A (en) Safety monitoring method and system based on big data analysis
CN115496144A (en) Power distribution network operation scene determining method and device, computer equipment and storage medium
CN116977663A (en) Image data processing method, device, equipment and medium
CN114971375A (en) Examination data processing method, device, equipment and medium based on artificial intelligence
CN112990789B (en) User health risk analysis system
CN110490061A (en) A kind of uncertainties model and measure of characteristics of remote sensing image
CN116758432A (en) Natural geological disaster classification and identification method and system based on improved Resnet neural network
CN113052063A (en) Confidence threshold selection method, device, equipment and storage medium
CN116796729A (en) Text recommendation method, device, equipment and storage medium based on feature enhancement
CN113313323A (en) Method, system, electronic device and medium for predicting human output of research and development project
Dharmawan et al. Tsunami tide prediction in shallow water using recurrent neural networks: Model implementation in the Indonesia Tsunami Early Warning System
CN113157755A (en) User data processing method, system, electronic device and medium
Zeng et al. Adaptive deep neural networks for solving corner singular problems
Ivanov et al. Use of Bootstrap Interface for Implementation of the Algorithm of a Small Sample Increase of Biometric Data
CN115083442B (en) Data processing method, device, electronic equipment and computer readable storage medium
Sinha et al. Well-test model identification with self-organizing feature map
Ziegler et al. Application of sensitivities on simulation-data-based metamodels during the product synthesis
Wang et al. An Adaptive Logarithm Quantization Method for DNN Compression
CN115034396A (en) Model training method and device, storage medium and electronic equipment

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
GR01 Patent grant
GR01 Patent grant
TR01 Transfer of patent right
TR01 Transfer of patent right

Effective date of registration: 20220721

Address after: 201615 room 1904, G60 Kechuang building, No. 650, Xinzhuan Road, Songjiang District, Shanghai

Patentee after: Shanghai Mingping Medical Data Technology Co.,Ltd.

Address before: 102400 no.86-n3557, Wanxing Road, Changyang, Fangshan District, Beijing

Patentee before: Mingpinyun (Beijing) data Technology Co.,Ltd.