Intelligent physical examination data dynamic analysis method

Abstract

The invention relates to the technical field of data analysis, and discloses an intelligent physical examination data dynamic analysis method, which is characterized by comprising the following steps of: fixing a centroid matrix F (v) of vth test index data, a common clustering index matrix G and a weight factor alpha (v) of the vth test index data, and updating a diagonal matrix D (v) corresponding to a vth view; fixing G, D (v) and alpha (v) and updating the cluster centroid of each view F (v); fixing F (v), D (v) and alpha (v) and updating G; f (v), G and D (v) are fixed, alpha (v) is updated, potential models of data are mined from features of known training data, an effective model is formed by clustering each data to different clusters, then unknown new data are divided by using the model, physical state clustering and evaluated multi-view clustering of a tested person are carried out, and the optimal physical state of the tested person is obtained. And physical examination test indexes are taken as responses of physical conditions from different data perspectives.

Description

Technical Field

[0001] The present invention relates to the technical field of data analysis, and more specifically, to a method for dynamic analysis of intelligent physical examination data. Background Art

[0002] Due to the limitations of equipment, time, climate and other factors, it is very difficult to conduct physical examinations of all 15 items on a regular basis. Moreover, it relies on the experience of the testers and needs to judge the physical condition of the person being tested during the test preparation process. This requires the testers to have rich physical examination experience, which increases the pressure on the testers and the possibility of error. Therefore, with the accumulation of physical examination data, traditional physical examination data processing and analysis methods can no longer meet actual needs.

[0003] Therefore, the present invention provides an intelligent physical examination data dynamic analysis method to improve the above technical problems. Summary of the invention

[0004] The disclosed embodiments aim to address the deficiencies of the prior art and provide an intelligent physical examination data dynamic analysis method. The present invention adopts a multi-view clustering method to explore the importance of each physical examination test indicator and perform clustering evaluation on the physical condition of the person being tested.

[0005] The above technical objectives of the present invention are achieved through the following technical solutions: A method for dynamic analysis of intelligent physical examination data, comprising the following steps:

[0006] S1. Normalize the data of each physical examination test indicator;

[0007] S2, fix the centroid matrix F of the υth test index data (υ) , public clustering index matrix G, weight factor α of the υth test index data (υ) , update the diagonal matrix D corresponding to the υth view (υ) ;

[0008] S3, fixed G, D (υ) , α (υ) And update each view F (υ) The cluster centroid of

[0009] S4, Fixed F (υ) , D (υ) , α (υ) And update G;

[0010] S5, Fixed F (υ) , G, D (υ) And update α (υ) .

[0011] As a preferred technical solution of the present invention, is the υth test index data matrix of the physical examination data, is the clustering index matrix of the υth test index data, is the centroid matrix of the υth test index data, is the public clustering indicator matrix;

[0012] Since many different test index data indicate the physical condition of the same person being tested, the clustering index matrix G of different test index data is (υ) is unique; using the structured sparsity induced norm L 2，1 , this norm is robust to data outliers and obtains stable results with different initializations;

[0013] The loss function can be expressed as:

[0014]

[0015] Among them, H (υ) =Tr{(X (υ) -F (υ) G T )D (υ) (X (υ) -F (υ) G T ) T} (2)

[0016] In the formula, α (υ) is the weight factor of the υth test index data, γ is the parameter controlling the weight distribution of different test index data, υ = 1, 2, …, M, M is M kinds of heterogeneous features.

[0017] As a preferred technical solution of the present invention, update the diagonal matrix D corresponding to the vth view (υ) The process is: is the diagonal matrix corresponding to the υth view, and the i-th entry on the diagonal is defined as:

[0018]

[0019] where e (υ)i is the i-th row of the following matrix:

[0020]

[0021] As a preferred technical solution of the present invention, update each view F (υ) The process of cluster centroid is: for J, for F (υ) Taking the derivative, we get:

[0022]

[0023] in

[0024] When equation (5) is set to 0, F can be updated (υ) :

[0025]

[0026] As a preferred technical solution of the present invention, the process of updating the clustering index matrix G is:

[0027]

[0028] For a fixed specific i, the vector

[0029]

[0030] g * =e k (10)

[0031] in, is a matrix The i-th element on the diagonal; k is determined as follows:

[0032]

[0033] As a preferred technical solution of the present invention, update α (υ) The process is:

[0034]

[0035] Solve formula (12) through Lagrangian function:

[0036]

[0037] To obtain the optimal solution, set the derivative α of equation (13) (υ) is zero:

[0038]

[0039] The obtained α (υ) Substituting into equation (14), let For the constraints:

[0040]

[0041] In summary, the present invention has the following beneficial effects: by mining the potential model of the data from the features of known training data, by clustering each data into different clusters with an effective description or model, and then using the model to divide unknown new data, clustering the physical state of the person being tested and the multi-view clustering of the evaluation, and using the physical examination test indicators as a reflection of the physical condition from different data perspectives. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 The present invention provides a flowchart of a method for dynamic analysis of intelligent physical examination data. DETAILED DESCRIPTION

[0043] The present application is described in detail below in conjunction with specific embodiments. The following embodiments will help those skilled in the art to further understand the present application, but do not limit the present application in any form. It should be noted that, for those of ordinary skill in the art, several variations and improvements can also be made without departing from the concept of the present application. These all belong to the protection scope of the present application.

[0044] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0045] It should be noted that, if there is no conflict, the various features in the embodiments of the present application can be combined with each other, all within the scope of protection of the present application. In addition, although the functional module division is performed in the device schematic diagram and the logical order is shown in the flow chart, in some cases, the steps shown or described can be performed in a sequence different from the module division in the device or the flow chart. In addition, the words "first", "second", "third", etc. used herein do not limit the data and execution order, but only distinguish the same items or similar items with basically the same functions and effects.

[0046] Unless otherwise defined, all technical and scientific terms used in this specification have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used in this specification and in the specification of this application are only for the purpose of describing specific embodiments and are not intended to limit this application. The term "and / or" used in this specification includes any and all combinations of one or more of the related listed items.

[0047] In addition, the technical features involved in each embodiment of the present application described below can be combined with each other as long as they do not conflict with each other.

[0048] The disclosed embodiment aims to solve the problem that it is very difficult to conduct physical examination tests for all 15 items regularly; it requires the test personnel to have rich physical examination experience, which increases the pressure on the test personnel and the possibility of error. In view of this, the disclosed embodiment proposes an intelligent physical examination data dynamic analysis method that uses a multi-view clustering method to mine the importance of each physical examination test indicator and perform cluster evaluation on the physical condition of the tested person.

[0049] Please refer to Figure 1 , Figure 1 The flowchart of a method for dynamic analysis of intelligent physical examination data according to an embodiment of the present disclosure is shown. The overall process mainly includes the following five steps:

[0050] S1. Normalize each physical examination test indicator data so that the value of each physical examination test indicator data is within the range of [-1; 1].

[0051] Connect all standardized data matrices as K-means input; use the parameter γ to control the weight distribution between all physical examination test indicator data. Use K-fold cross validation to select the optimal number of clusters.

[0052] Specifically, is the υth test index data matrix of the physical examination data, is the clustering index matrix of the υth test index data, is the centroid matrix of the υth test index data, is the public clustering indicator matrix.

[0053] Since many different test index data indicate the physical condition of the same person being tested, the clustering index matrix G of different test index data is (υ) Should be unique. Use the structured sparsity induced norm L 2，1 , which is robust to data outliers and obtains stable results with different initializations. The loss function can be expressed as:

[0054]

[0055] Among them, H (υ) =Tr{X (υ) F (υ) G T )D (υ) (X (υ) -F (υ) G T ) T} (2)

[0056] In the formula, α (υ)is the weight factor of the υth test index data, γ is the parameter controlling the weight distribution of different test index data, υ = 1, 2, …, M, M is M kinds of heterogeneous features.

[0057] S2, fixed F (υ) , G, α (υ) And update D by formulas (3) and (4) (υ) .

[0058] is the diagonal matrix corresponding to the υth view, and the i-th entry on the diagonal is defined as:

[0059]

[0060] where e (υ)i is the i-th row of the following matrix:

[0061]

[0062] S3, fixed G, D (υ) , α (υ) And update each view F (υ) The cluster centroid of .

[0063] For J and F (υ) Derivation, get

[0064]

[0065] in

[0066] When equation (5) is set to 0, F can be updated (υ) :

[0067]

[0068] S4, Fixed F (υ) , D (υ) , α (υ) And update the clustering indicator matrix G.

[0069]

[0070] For a fixed specific i, the vector

[0071]

[0072] g * =e k (10)

[0073] in, is a matrix The i-th element on the diagonal; the number of k clusters is determined as follows:

[0074]

[0075] S5, Fixed F (υ) , G, D (υ) And update α (υ) .

[0076]

[0077] Solve formula (12) through Lagrangian function:

[0078]

[0079] To obtain the optimal solution, set the derivative α of equation (13) (υ) is zero:

[0080]

[0081] The obtained α (υ) Substituting into equation (14), let For the constraints:

[0082]

[0083] Through the above four steps, F is updated alternately. (υ) , G, D (υ) and α (υ) ,The process is repeated iteratively until the objective function converges, realizing multi-view clustering and evaluation of the physical states of the tested persons, and using the physical examination test indicators as a reflection of the physical condition from different data perspectives.

[0084] The above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technicians in this technical field, some improvements and modifications without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.

Claims

1. A method for dynamic analysis of intelligent physical examination data, characterized in that: The method comprises the following steps: S1. Normalize the data of each physical examination test indicator; S2, fix the centroid matrix F of the vth test index data (v) , public clustering index matrix G, weight factor α of the vth test index data (v) , update the diagonal matrix D corresponding to the vth view (v) ; S3, fixed G, D (v) , α (v) And update each view F (v) The cluster centroid of S4, Fixed F (v) , D (v) , α (v) And update G; S5, Fixed F (v) , G, D (v) And update α (v) .

2. According to claim 1, a method for dynamic analysis of intelligent physical examination data is characterized in that: set up is the vth test index data matrix of the physical examination data, is the clustering index matrix of the vth test index data, is the centroid matrix of the vth test index data, is the public clustering indicator matrix; Since many different test index data indicate the physical condition of the same person being tested, the clustering index matrix G of different test index data is (v) is unique; Using the structured sparsity induced norm L 2，1 , this norm is robust to data outliers and obtains stable results with different initializations; The loss function can be expressed as: Among them, H (v) =Tr{(X (v) -F (v) G T )D (v) (X (v) -F (v) G T ) T } (2) In the formula, α (v) is the weight factor of the vth test index data, γ is the parameter controlling the weight distribution of different test index data, v = 1, 2, …, M, M is M kinds of heterogeneous features.

3. The intelligent physical examination data dynamic analysis method according to claim 1, characterized in that: Update the diagonal matrix D corresponding to the vth view (v) The process is: is the diagonal matrix corresponding to the v-th view, and the i-th entry on the diagonal is defined as: where e (v)i is the i-th row of the following matrix: F (v) =X (v)T -GF (v)T (4)。 4. The intelligent physical examination data dynamic analysis method according to claim 1, characterized in that: Update each view F (v) The process of cluster centroid is: for J, for F (v) Taking the derivative, we get: in: When equation (5) is set to 0, F can be updated (v) :

5. The intelligent physical examination data dynamic analysis method according to claim 1, characterized in that: The process of updating the clustering indicator matrix G is: For a fixed specific i, the vector g * =e k (10) in, is a matrix The i-th element on the diagonal; k is determined as follows:

6. The intelligent physical examination data dynamic analysis method according to claim 1, characterized in that: Update alpha (v) The process is: Solve formula (12) through Lagrangian function: To obtain the optimal solution, set the derivative α of equation (13) (v) is zero: The obtained α (v) Substituting into equation (14), let For the constraints:

Images