CN110097126A - The method that verification emphasis personnel based on DBSCAN clustering algorithm, house fail to register note - Google Patents
The method that verification emphasis personnel based on DBSCAN clustering algorithm, house fail to register note Download PDFInfo
- Publication number
- CN110097126A CN110097126A CN201910374115.XA CN201910374115A CN110097126A CN 110097126 A CN110097126 A CN 110097126A CN 201910374115 A CN201910374115 A CN 201910374115A CN 110097126 A CN110097126 A CN 110097126A
- Authority
- CN
- China
- Prior art keywords
- house
- emphasis personnel
- sample
- feature
- personnel
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
-
- 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
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
Abstract
The method that the verification emphasis personnel that the present invention relates to a kind of based on DBSCAN clustering algorithm, house fail to register note, pre-processes population, the house data set of people's police's acquisition, including Missing Data Filling, classification type variable discretization, numeric type variable standardization；Using DBSCAN clustering algorithm, the sample of non-core point is sorted out on " emphasis personnel, house " data set, cluster result is analyzed；The data of all labels " emphasis personnel, house " are carried out core point to fix, pass through the DBSCAN clustering algorithm of self-adaptive features weight, the sample of non-core point is sorted out on population, house data set, obtain cluster result, ultimately generate it is doubtful fail to register note " emphasis personnel, house " verification table.Thus, personnel, house are paid close attention to as core using labeled, personnel similar with emphasis personnel, house, house are filtered out by density clustering algorithm, then reduce doubtful emphasis personnel, house verification range, the precision and efficiency of police service verification can be effectively improved.
Description
Technical field
The present invention relates to a kind of method that verification emphasis personnel, house fail to register note, more particularly to one kind are poly- based on DBSCAN
The method that the verification emphasis personnel of class algorithm, house fail to register note.
Background technique
In big data era, data mining technology has played huge effect in many fields.Pass through big data and algorithm
Optimization traditional community's police service verifies operating mode to improve a series of problems of public security basic-level policemen as received basis state.It is initially cut
Field is exactly police field, public security data not only magnanimity, and type is abundant, existing traditional structural data, also has a large amount of
Unstructured data.Population, building management work are heavy, and traditional police service verifies working method and has been difficult to meet so big
Population, the public security population under house radix, house management.
It is needed for the investigation in emphasis personnel in public security data, house, divides and hierarchical method is relatively early to propose to be also ratio
Relatively effective basic clustering method, such as K-means, K-modes, but they are intended to find spherical cluster, are but difficult discovery and appoint
The cluster for shape of anticipating, and algorithm needs to preset K value.Density clustering such as DBSCAN can regard cluster as quilt in data space
The separated dense Region in sparse region has the advantages that any cluster of discovery, excludes noise spot automatically, need not specify classification number.
But the core point that classical density clustering algorithm obtains contains all the points for meeting Neighbor Condition, population,
In house clustering problem, as core point, Clustering Effect is unsatisfactory for non-emphasis personnel, house, furthermore different characteristic away from
The characteristic attribute that same weight can weaken those and emphasis personnel, house strong correlation is assigned during from similarity is calculated, from
And it will appear wrong poly- situation.
In view of the above shortcomings, the designer, is actively subject to research and innovation, it is a kind of based on DBSCAN cluster to found
The method that the verification emphasis personnel of algorithm, house fail to register note makes it with more the utility value in industry.
Summary of the invention
In order to solve the above technical problems, the object of the present invention is to provide a kind of verification emphasis based on DBSCAN clustering algorithm
The method that personnel, house fail to register note.
The method that verification emphasis personnel based on DBSCAN clustering algorithm of the invention, house fail to register note, it is characterised in that
The following steps are included:
Step 1 pre-processes population, the house data set of people's police's acquisition, including Missing Data Filling, classification type become
Measure discretization, numeric type variable standardization；
Step 2 is known " emphasis personnel, house " and unknown by being divided by the data of label " emphasis personnel, house "
" emphasis personnel, house ", and " emphasis personnel, house " data sample by known to is fixed as the core point of Density Clustering, and separates
Non-core point；
Step 3 is set Neighbourhood parameter (ε, MinPts), and ε describes the neighborhood distance threshold of a certain sample here,
MinPts describes the distance of a certain sample as the threshold value of number of samples in the neighborhood of ε.Using DBSCAN clustering algorithm, in " weight
The sample of non-core point is sorted out on point personnel, house " data set, cluster result is analyzed；
The data of all labels " emphasis personnel, house " are carried out core point and fixed, weighed by self-adaptive features by step 4
The DBSCAN clustering algorithm of weight, sorts out the sample of non-core point on population, house data set, obtains cluster result；
Step 5 is counted and is judged to cluster result in step 4, ultimately generate it is doubtful fail to register note " emphasis personnel,
Table is verified in house ".
Further, the method that the above-mentioned verification emphasis personnel based on DBSCAN clustering algorithm, house fail to register note,
In, in step 1, the data prediction step is special to the population in public security population, house database, house related data
Sign is pre-processed, and including carrying out one-hot coding to the classification type feature in population, house related data feature, logarithm type is special
It levies variable and carries out nondimensionalization processing, the Missing Data Filling is to be filled to classification type feature with mode, and logarithm type feature becomes
Amount is filled with average.
Further, the method that the above-mentioned verification emphasis personnel based on DBSCAN clustering algorithm, house fail to register note,
In, the classification type feature includes gender, marital status, and the numeric type characteristic variable includes age, address longitude and latitude.
Further, the method that the above-mentioned verification emphasis personnel based on DBSCAN clustering algorithm, house fail to register note,
In, the processing of classification type variable discretization described in step 1 are as follows: assuming that there is N kind qualitative value, then it is special this feature to be extended to N kind
Sign, when primitive character value is i-th kind of qualitative value, i-th of extension feature is assigned a value of 1, other extension features are assigned a value of 0.It is described
The processing of numeric type variable standardization needs to calculate the mean value of every one-dimensional characteristicWith standard deviation (S), calculation formula is,
Further, the method that the above-mentioned verification emphasis personnel based on DBSCAN clustering algorithm, house fail to register note,
In, the process that the core point of Density Clustering is fixed as described in step 2 is to give " emphasis personnel, house " during calculating distance
Data sample assigns weight, and bigger positive value represents the sample and easily becomes core point, and smaller negative value can hinder sample to become
Core point.
Further, the method that the above-mentioned verification emphasis personnel based on DBSCAN clustering algorithm, house fail to register note,
In, in the step 3, by DBSCAN algorithm, the similarity between sample is measured using Euclidean distance, apart from smaller, sample
This is more similar, if n sample is divided into K cluster, the number of samples in each cluster is respectively as follows: n1, n2..., nk, then all K
The sum of inter- object distance of a cluster on jth dimensional feature dpFor,
xijFor the jth dimensional feature numerical value of i-th of sample, mkjFor mean value of the cluster k on jth dimensional feature, own
The sum of between class distance of the K cluster on jth dimensional feature dqFor,
mjFeature j is calculated to the contribution degree c of cluster later for mean value of the data set on jth dimensional featurej,
Finally, the feature weight w of jth dimensional featurejFor,
The dimension of m expression sample characteristics.
Euclidean distance formula to be weighted, so that the similarity d (m, n) between sample is obtained,
Further, the method that the above-mentioned verification emphasis personnel based on DBSCAN clustering algorithm, house fail to register note,
In, in the step 4, the treatment process fixed for core point be, using Scikit-learn machine learning frame, according to
Given Neighbourhood parameter finds out all core points.
Further, the method that the above-mentioned verification emphasis personnel based on DBSCAN clustering algorithm, house fail to register note,
In, in the step 4, feature weight is optimized, will be first divided by the data of label " emphasis personnel, house "
Know it is " emphasis personnel, house " and unknown " emphasis personnel, house ", according to core point fixing step, will known to " emphasis personnel, room
Room " data sample is fixed as the core point of Density Clustering, sets suitable Neighbourhood parameter, is then based on DBSCAN clustering algorithm,
The sample of non-core point is sorted out on " emphasis personnel, house " data set, each attribute pair is calculated for the result of classification
The contribution degree of cluster updates feature weight.
Still further, the method that the above-mentioned verification emphasis personnel based on DBSCAN clustering algorithm, house fail to register note,
In, it include to sentence by the quantity N of label " emphasis personnel, house " to being set in cluster result in each class in the step 5
Whether disconnected N is more than or equal to preset threshold value T, if it is judged that being N >=T, then not by the personnel of label, house in such
Note " emphasis personnel, house " are failed to register there are high likelihood is doubtful, ultimately generates and doubtful fails to register note " emphasis personnel, house " verification
Table；Otherwise, note " emphasis personnel, house " are failed to register there are low possibility is doubtful in such, needs to carry out artificial judgment.
According to the above aspect of the present invention, the present invention has at least the following advantages:
1, personnel, house are paid close attention to as core using labeled, is filtered out and emphasis personnel's phase by density clustering algorithm
As personnel, house, then reduce doubtful emphasis personnel, house verify range, can effectively improve police service verification precision and
Efficiency.
2, feature weight adaptation mechanism assigns different feature weights to attribute, can more accurately reflected sample it
Between similitude and improve clustering performance.
3, intelligent recommendation police service verifies mode, and there is verification object can be predicted, be prejudged in advance, verify work more section
It learns, precisely, people's police's duties more active, safety.
4, the upgrading of police service Model Transformation is pushed, is promoted and controls complicated population, house security administration ability aspect with important
Realistic function.
The above description is only an overview of the technical scheme of the present invention, in order to better understand the technical means of the present invention,
And can be implemented in accordance with the contents of the specification, the following is a detailed description of the preferred embodiments of the present invention and the accompanying drawings.
Detailed description of the invention
Fig. 1 is the flow diagram that feature weight optimizes.
Fig. 2 is the flow diagram of the DBSCAN clustering algorithm of self-adaptive features weight.
Specific embodiment
With reference to the accompanying drawings and examples, specific embodiments of the present invention will be described in further detail.Implement below
Example is not intended to limit the scope of the invention for illustrating the present invention.
The method that verification emphasis personnel based on DBSCAN clustering algorithm as shown in Figure 1, Figure 2, house fail to register note, with it is many not
Be with place the following steps are included:
Firstly, population, the house data set to people's police's acquisition pre-process, including Missing Data Filling, classification type variable
Discretization, numeric type variable standardization.Specifically, for the population in public security population, house database, house related data
Feature is pre-processed, including carrying out one-hot coding, logarithm type to the classification type feature in population, house related data feature
Characteristic variable carries out nondimensionalization processing.In order to meet the classification of facilitation, classification type feature includes gender, marital status, number
Value type characteristic variable includes age, address longitude and latitude.
The purpose for the arrangement is that removing, dimension is different to cause difference between data excessive, holds during calculating distance
Influence vulnerable to the larger feature of numerical value.The Missing Data Filling used is fills, logarithm type is special to classification type feature with mode
Sign variable is filled with average.The reason is that if the key feature comprising missing values, which directly abandons, to tie cluster
Fruit produces bigger effect.
The processing of the classification type variable discretization of use are as follows: assuming that there is N kind qualitative value, then this feature is extended in N
Feature, when primitive character value is i-th kind of qualitative value, i-th of extension feature is assigned a value of 1, other extension features are assigned a value of 0.
The processing of the numeric type variable standardization of use needs to calculate the mean value of every one-dimensional characteristicWith standard deviation (S), meter
Calculating formula is,
Then, it will be divided by the data of label " emphasis personnel, house " known " emphasis personnel, house " and unknown
" emphasis personnel, house ", and " emphasis personnel, house " data sample by known to is fixed as the core point of Density Clustering, and separates
Non-core point.
What the present invention used is fixed as the process of the core point of Density Clustering to give " emphasis people during calculating distance
Member, house " data sample assigns weight, and bigger positive value represents the sample and easily becomes core point, and smaller negative value can hinder
Sample becomes core point.
Later, Neighbourhood parameter (ε, MinPts) is set, ε describes the neighborhood distance threshold of a certain sample, MinPts here
The distance for describing a certain sample is the threshold value of number of samples in the neighborhood of ε.
Using DBSCAN clustering algorithm, the sample of non-core point is sorted out on " emphasis personnel, house " data set,
Cluster result is analyzed.
Specifically, by DBSCAN algorithm, the similarity between sample is measured using Euclidean distance, apart from smaller,
Sample is more similar, if n sample is divided into K cluster, the number of samples in each cluster is respectively as follows: n1, n2..., nk, then institute
There is the sum of inter- object distance of the K cluster on jth dimensional feature dpFor,
Wherein:
xijFor the jth dimensional feature numerical value of i-th of sample, mkjFor mean value of the cluster k on jth dimensional feature.
Meanwhile the sum of the between class distance of all K clusters on jth dimensional feature dqFor,
Wherein, mjFeature j is calculated to the contribution degree c of cluster later for mean value of the data set on jth dimensional featurej,
Finally, the feature weight w of jth dimensional featurejFor,
Wherein, m indicates the dimension of sample characteristics.
Euclidean distance formula to be weighted, so that the similarity d (m, n) between sample is obtained,
Then, the data of all labels " emphasis personnel, house " are carried out core point to fix, passes through self-adaptive features weight
DBSCAN clustering algorithm, the sample of non-core point is sorted out on population, house data set, obtain cluster result.
Specifically, the treatment process fixed for core point be, using Scikit-learn machine learning frame, according to
Given Neighbourhood parameter (ε, MinPts) finds out all core points.Also, feature weight is optimized, it first will be by label
The data of " emphasis personnel, house " be divided into it is known " emphasis personnel, house " and unknown " emphasis personnel, house ", according to core
Point fixing step, " emphasis personnel, house " data sample by known to are fixed as the core point of Density Clustering, set suitable neighborhood
Parameter.
Then, it is based on DBSCAN clustering algorithm, the sample of non-core point is carried out on " emphasis personnel, house " data set
Sort out, each attribute is calculated to the contribution degree of cluster for the result of classification, updates feature weight.
Finally, counted and judged to cluster result, ultimately generates and doubtful fail to register note " emphasis personnel, house " verification
Table.It specifically, include to judge N by the quantity N of label " emphasis personnel, house " to being set in cluster result in each class
Whether preset threshold value T is more than or equal to.
Specifically, if it is judged that being N >=T, then by the personnel of label, house, there are high likelihoods to doubt in such
Seemingly fail to register note " emphasis personnel, house ", ultimately generate it is doubtful fail to register note " emphasis personnel, house " verify table.Otherwise, it is deposited in such
Note " emphasis personnel, house " are failed to register low possibility is doubtful, needs to carry out artificial judgment, expert is needed by virtue of experience to judge again,
Reduce artificial erroneous judgement.Further, it is possible to push verification task to people's police.
The present invention considers personnel in public security population management work, house complexity, the feature of people's police's lazy weight as a result,
By the screening and clustering to data, chosen on having fully considered feature base possessed by " emphasis personnel, house "
Cluster feature appropriate fails to register note phenomenon to " emphasis personnel, house " and has carried out more accurate judgement.Meanwhile it generating doubtful
Note verification table is failed to register, pushes verification task in terms of facilitating public security.By density clustering algorithm diminution fail to register note pay close attention to personnel,
Range is verified in house, is finally improved population, the precision that house is verified, is also provided decision support for the other control fields of public security.
Working principle of the present invention is as follows:
From the point of view of table 1, before population (house is similar) data prediction:
| Name number | Age | Gender | Schooling | Latitude | Longitude |
| 1 | 73 | 2 | Null | 31.323771 | 120.666739 |
| 2 | 53 | 1 | Null | 31.315803 | 120.665558 |
| 3 | 46 | 2 | Null | 31.317036 | 120.747582 |
| 4 | 29 | 2 | 70 | 34.646452 | 116.912783 |
| 5 | 21 | 1 | 40 | 32.066899 | 118.193343 |
| 6 | 46 | 1 | 70 | 27.221181 | 111.248449 |
| 7 | 44 | 1 | 20 | 31.319655 | 120.731328 |
| 8 | 62 | 2 | Null | 31.320779 | 120.665973 |
| 9 | 31 | 1 | 60 | 35.828924 | 116.013732 |
| 10 | 32 | 1 | 60 | 34.357221 | 115.363676 |
Wherein:
Gender: 1- male, 2- female；Schooling: 20- undergraduate course, 40- special secondary school, 60- senior middle school, the junior middle school 70-, Null- missing.
After being handled using method of the invention, it is as follows to obtain table 2:
The demonstration data preprocessing process by taking name number 1 as an example:
Classification type variable discretization: sex character has 2 kinds of qualitative values, respectively male, female, then this feature is extended to 2
Kind feature, at this point, the sex character value of name number 1 is the 2nd kind of qualitative value, so the 2nd extension feature is assigned a value of 1, other
Extension feature is assigned a value of 0, uses (0,1) to indicate sex character after discretization.
Numeric type variable standardization: firstly the need of calculate age characteristics mean value and standard deviation, respectively 43.7,
15.23187447, age value is after being standardized according to standardized calculation formula,
(73-43.7)/15.23187447=1.923598.
Longitude, latitude standardized value can similarly be acquired.
Missing Data Filling processing: schooling includes missing values Null, is filled using the mode 60 of existing classification, is then adopted
With sliding-model control mode same as sex character.
It can be seen from the above written description and the attached drawings that after applying the present invention, gathering around and having the following advantages:
1, personnel, house are paid close attention to as core using labeled, is filtered out and emphasis personnel, room by density clustering algorithm
The similar personnel in room, house, then doubtful emphasis personnel, house verification range are reduced, the accurate of police service verification can be effectively improved
Degree and efficiency.
2, feature weight adaptation mechanism assigns different feature weights to attribute, can more accurately reflected sample it
Between similitude and improve clustering performance.
3, intelligent recommendation police service verifies mode, and there is verification object can be predicted, be prejudged in advance, verify work more section
It learns, precisely, people's police's duties more active, safety.
4, the upgrading of police service Model Transformation is pushed, is promoted and controls complicated population, house security administration ability aspect with important
Realistic function.
The above is only a preferred embodiment of the present invention, it is not intended to restrict the invention, it is noted that for this skill
For the those of ordinary skill in art field, without departing from the technical principles of the invention, can also make it is several improvement and
Modification, these improvements and modifications also should be regarded as protection scope of the present invention.
Claims (9)
1. the method that the verification emphasis personnel based on DBSCAN clustering algorithm, house fail to register note, it is characterised in that including following step
It is rapid:
Step 1, to people's police acquisition population, house data set pre-process, including Missing Data Filling, classification type variable from
Dispersion, numeric type variable standardization；
Step 2 will be divided into known " emphasis personnel, house " and unknown " weight by the data of label " emphasis personnel, house "
Point personnel, house ", and " emphasis personnel, house " data sample by known to is fixed as the core point of Density Clustering, and separates non-core
Heart point；
Step 3 is set Neighbourhood parameter (ε, MinPts), and ε describes the neighborhood distance threshold of a certain sample here, and MinPts is retouched
The distance for having stated a certain sample is the threshold value of number of samples in the neighborhood of ε.Using DBSCAN clustering algorithm, in " emphasis personnel, room
The sample of non-core point is sorted out on room " data set, cluster result is analyzed；
The data of all labels " emphasis personnel, house " are carried out core point and fixed, pass through self-adaptive features weight by step 4
DBSCAN clustering algorithm sorts out the sample of non-core point on population, house data set, obtains cluster result；
Step 5 is counted and is judged to cluster result in step 4, is ultimately generated and doubtful is failed to register note " emphasis personnel, house "
Verify table.
2. the method that the verification emphasis personnel according to claim 1 based on DBSCAN clustering algorithm, house fail to register note,
Be characterized in that: in step 1, the data prediction step is related to population, the house in public security population, house database
Data characteristics is pre-processed, including carrying out one-hot coding, logarithm to the classification type feature in population, house related data feature
Value type characteristic variable carries out nondimensionalization processing, and the Missing Data Filling is to be filled to classification type feature with mode, logarithm type
Characteristic variable is filled with average.
3. the method that the verification emphasis personnel according to claim 1 based on DBSCAN clustering algorithm, house fail to register note,
Be characterized in that: the classification type feature includes gender, marital status, and the numeric type characteristic variable includes age, address longitude and latitude
Degree.
4. the method that the verification emphasis personnel according to claim 1 based on DBSCAN clustering algorithm, house fail to register note,
It is characterized in that: the processing of classification type variable discretization described in step 1 are as follows: assuming that there is N kind qualitative value, then extend this feature
For N kind feature, when primitive character value is i-th kind of qualitative value, i-th of extension feature is assigned a value of 1, other extension features are assigned a value of
0.The processing of the numeric type variable standardization needs to calculate the mean value of every one-dimensional characteristicWith standard deviation (S), calculation formula
For,
5. the method according to claim 1 that verify emphasis personnel, house registration based on DBSCAN clustering algorithm, special
Sign is: the process that the core point of Density Clustering is fixed as described in step 2 is, during calculating distance to " emphasis personnel,
House " data sample assigns weight, and bigger positive value represents the sample and easily becomes core point, and smaller negative value can hinder sample
As core point.
6. the method that the verification emphasis personnel according to claim 1 based on DBSCAN clustering algorithm, house fail to register note,
It is characterized in that: in the step 3, by DBSCAN algorithm, the similarity between sample, distance is measured using Euclidean distance
Smaller, sample is more similar, if n sample is divided into K cluster, the number of samples in each cluster is respectively as follows: n1, n2..., nk,
Then the sum of the inter- object distance of all K clusters on jth dimensional feature dpFor,
xijFor the jth dimensional feature numerical value of i-th of sample, mkjFor mean value of the cluster k on jth dimensional feature, own
The sum of between class distance of the K cluster on jth dimensional feature dqFor,
mjFeature j is calculated to the contribution degree c of cluster later for mean value of the data set on jth dimensional featurej,
Finally, the feature weight w of jth dimensional featurejFor,
The dimension of m expression sample characteristics.
Euclidean distance formula to be weighted, so that the similarity d (m, n) between sample is obtained,
7. the method that the verification emphasis personnel according to claim 1 based on DBSCAN clustering algorithm, house fail to register note,
It is characterized in that: in the step 4, being for the fixed treatment process of core point, using Scikit-learn machine learning frame
Frame finds out all core points according to given Neighbourhood parameter.
8. the method that the verification emphasis personnel according to claim 1 based on DBSCAN clustering algorithm, house fail to register note,
It is characterized in that: in the step 4, feature weight being optimized, will first be drawn by the data of label " emphasis personnel, house "
It is " emphasis personnel, house " and unknown " emphasis personnel, house " known to being divided into, according to core point fixing step, " the emphasis people by known to
Member, house " data sample is fixed as the core point of Density Clustering, sets suitable Neighbourhood parameter, is then based on DBSCAN cluster
Algorithm sorts out the sample of non-core point on " emphasis personnel, house " data set, calculates each category for the result of classification
Property to the contribution degree of cluster, update feature weight.
9. the method that the verification emphasis personnel according to claim 1 based on DBSCAN clustering algorithm, house fail to register note,
It is characterized in that: including by the number of label " emphasis personnel, house " to being set in cluster result in each class in the step 5
N is measured, judges whether N is more than or equal to preset threshold value T, if it is judged that being N >=T, then not by the people of label in such
Member, house fail to register note " emphasis personnel, house " there are high likelihood is doubtful, ultimately generate and doubtful fail to register note " emphasis personnel, room
Table is verified in room "；Otherwise, note " emphasis personnel, house " are failed to register there are low possibility is doubtful in such, needs to carry out artificial judgment.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910374115.XA CN110097126B (en) | 2019-05-07 | 2019-05-07 | Method for checking important personnel and house missing registration based on DBSCAN clustering algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910374115.XA CN110097126B (en) | 2019-05-07 | 2019-05-07 | Method for checking important personnel and house missing registration based on DBSCAN clustering algorithm |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110097126A true CN110097126A (en) | 2019-08-06 |
| CN110097126B CN110097126B (en) | 2023-04-21 |
Family
ID=67447081
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910374115.XA Active CN110097126B (en) | 2019-05-07 | 2019-05-07 | Method for checking important personnel and house missing registration based on DBSCAN clustering algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110097126B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110650145A (en) * | 2019-09-26 | 2020-01-03 | 湖南大学 | Low-rate denial of service attack detection method based on SA-DBSCAN algorithm |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103902655A (en) * | 2014-02-28 | 2014-07-02 | 小米科技有限责任公司 | Clustering method and device and terminal device |
| CN103902654A (en) * | 2014-02-28 | 2014-07-02 | 小米科技有限责任公司 | Clustering method and device and terminal device |
| CN106600059A (en) * | 2016-12-13 | 2017-04-26 | 北京邮电大学 | Intelligent power grid short-term load predication method based on improved RBF neural network |
| CN107993179A (en) * | 2018-01-04 | 2018-05-04 | 南京市公安局栖霞分局 | A kind of police service platform population house data examination register method |
| CN108280479A (en) * | 2018-01-25 | 2018-07-13 | 重庆大学 | A kind of power grid user sorting technique based on Load characteristics index weighted cluster algorithm |
| CN108875806A (en) * | 2018-05-31 | 2018-11-23 | 中南林业科技大学 | False forest fires hot spot method for digging based on space-time data |
-
2019
- 2019-05-07 CN CN201910374115.XA patent/CN110097126B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103902655A (en) * | 2014-02-28 | 2014-07-02 | 小米科技有限责任公司 | Clustering method and device and terminal device |
| CN103902654A (en) * | 2014-02-28 | 2014-07-02 | 小米科技有限责任公司 | Clustering method and device and terminal device |
| CN106600059A (en) * | 2016-12-13 | 2017-04-26 | 北京邮电大学 | Intelligent power grid short-term load predication method based on improved RBF neural network |
| CN107993179A (en) * | 2018-01-04 | 2018-05-04 | 南京市公安局栖霞分局 | A kind of police service platform population house data examination register method |
| CN108280479A (en) * | 2018-01-25 | 2018-07-13 | 重庆大学 | A kind of power grid user sorting technique based on Load characteristics index weighted cluster algorithm |
| CN108875806A (en) * | 2018-05-31 | 2018-11-23 | 中南林业科技大学 | False forest fires hot spot method for digging based on space-time data |
Non-Patent Citations (2)
| Title |
|---|
| 付家祺 等: ""一种基于密度和约束的数据流聚类算法"", 《科技创新与应用》 * |
| 李轶 等: ""一种 DBSCAN 聚类点密度的加权质心定位算法"", 《河南科技大学学报（自然科学版）》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110650145A (en) * | 2019-09-26 | 2020-01-03 | 湖南大学 | Low-rate denial of service attack detection method based on SA-DBSCAN algorithm |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110097126B (en) | 2023-04-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103927302B (en) | A kind of file classification method and system | |
| CN106408030B (en) | SAR image classification method based on middle layer semantic attribute and convolutional neural networks | |
| CN105045875B (en) | Personalized search and device | |
| CN110647904B (en) | Cross-modal retrieval method and system based on unmarked data migration | |
| CN111881290A (en) | Distribution network multi-source grid entity fusion method based on weighted semantic similarity | |
| WO2022126810A1 (en) | Text clustering method | |
| CN107122352A (en) | A kind of method of the extracting keywords based on K MEANS, WORD2VEC | |
| CN110209808A (en) | A kind of event generation method and relevant apparatus based on text information | |
| CN114841257B (en) | Small sample target detection method based on self-supervision comparison constraint | |
| CN110147421B (en) | Target entity linking method, device, equipment and storage medium | |
| CN112685504B (en) | Production process-oriented distributed migration chart learning method | |
| CN108710894A (en) | A kind of Active Learning mask method and device based on cluster representative point | |
| CN108470022A (en) | A kind of intelligent work order quality detecting method based on operation management | |
| CN106326923B (en) | A kind of position data clustering method of registering taking position into account and repeating with density peaks point | |
| CN109063113A (en) | A kind of fast image retrieval method based on the discrete Hash of asymmetric depth, retrieval model and model building method | |
| CN111177402B (en) | Evaluation method, device, computer equipment and storage medium based on word segmentation processing | |
| CN107145516A (en) | A kind of Text Clustering Method and system | |
| CN109657147A (en) | Microblogging abnormal user detection method based on firefly and weighting extreme learning machine | |
| CN104699817B (en) | A kind of method for sequencing search engines and system based on improvement spectral clustering | |
| CN107977670A (en) | Accident classification stage division, the apparatus and system of decision tree and bayesian algorithm | |
| CN111539612B (en) | Training method and system of risk classification model | |
| Wilkins et al. | Comparison of five clustering algorithms to classify phytoplankton from flow cytometry data | |
| CN115114409A (en) | Civil aviation unsafe event combined extraction method based on soft parameter sharing | |
| CN110097126A (en) | The method that verification emphasis personnel based on DBSCAN clustering algorithm, house fail to register note | |
| CN106775694B (en) | A kind of hierarchy classification method of software configuration code product |
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 |