CN105678428A - Criminal suspicion probability prediction method and system - Google Patents
Criminal suspicion probability prediction method and system Download PDFInfo
- Publication number
- CN105678428A CN105678428A CN201610084782.0A CN201610084782A CN105678428A CN 105678428 A CN105678428 A CN 105678428A CN 201610084782 A CN201610084782 A CN 201610084782A CN 105678428 A CN105678428 A CN 105678428A
- Authority
- CN
- China
- Prior art keywords
- previous conviction
- type
- assignment
- conviction
- previous
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 36
- 238000001514 detection method Methods 0.000 claims abstract description 38
- 238000012549 training Methods 0.000 claims abstract description 28
- 238000007477 logistic regression Methods 0.000 claims abstract description 10
- 238000012216 screening Methods 0.000 claims description 13
- 230000015572 biosynthetic process Effects 0.000 claims description 6
- 238000007689 inspection Methods 0.000 abstract 1
- 238000004458 analytical method Methods 0.000 description 5
- 230000006870 function Effects 0.000 description 4
- 238000011835 investigation Methods 0.000 description 4
- 238000006243 chemical reaction Methods 0.000 description 2
- 230000001419 dependent effect Effects 0.000 description 2
- 238000013024 troubleshooting Methods 0.000 description 2
- PEDCQBHIVMGVHV-UHFFFAOYSA-N Glycerine Chemical compound OCC(O)CO PEDCQBHIVMGVHV-UHFFFAOYSA-N 0.000 description 1
- 238000009825 accumulation Methods 0.000 description 1
- 230000003321 amplification Effects 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000013277 forecasting method Methods 0.000 description 1
- 238000003199 nucleic acid amplification method Methods 0.000 description 1
- 238000000611 regression analysis Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 208000011580 syndromic disease Diseases 0.000 description 1
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
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- 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—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Tourism & Hospitality (AREA)
- Human Resources & Organizations (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Marketing (AREA)
- Development Economics (AREA)
- General Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Quality & Reliability (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Educational Administration (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a criminal suspicion probability prediction method, and the method comprises the steps: obtaining the related information of a to-be-detected person, determining the corresponding criminal record types of the to-be-detected person according to the related information, and also determining the related historical data screened through a specified index corresponding to each criminal record type; determining the detection time, and carrying out the numerical value assignment of the specified indexes of all criminal record types according to the related historical data; setting a characteristic vector which is formed by all the specified indexes, and obtaining the training sample library of each criminal record type according to the characteristic vector and the values of all specified indexes; enabling the classification attribute of the training sample library of each criminal record type to be divided into 1 and 0, carrying out the fitting of the classification probability of the classification attribute (1) in the training sample library corresponding to each criminal record type through employing a logistic regression model, and obtaining the crime probability of all criminal record types of the to-be-detected person. According to the invention, the method can accurately detect the crime type and probability of the to-be-detected person, and provides onsite guide for public security officers to carry out key inspection.
Description
Technical field
The present invention relates to public security prediction of criminality technical field, particularly relate to the method and system of a kind of suspicion of crime probabilistic forecasting.
Background technology
At present, China is in critical period that Police Informationization and " information dominate police service " are deeply developed, and in the mode of operation of whole " information dominates police service ", the analysis of crime information and study and judge as core, prediction of criminality is then the most important thing. Generally, the method for excess syndrome is mainly taked in the research of prediction of criminality, by inquiry, data collection, analysis, conclusion, draw important correlation factor, thus disclose crime occur rule.
In prior art, domestic prediction of criminality mainly adopts the method for mathematical analysis, including regression analysis, gray system theory analytic process and optimum combination analytic process etc., but these methods all majorities lay particular stress on macro forecasting field, fail to lay particular stress on Mi-crocosmic forecast field, the probability size of criminal type and correspondence thereof cannot be provided, thus people's police's routine work is lacked good assosting effect for the information of concrete personage and movement track.
Therefore, need badly a kind of suspicion of crime probabilistic forecasting method and, it is possible to being specifically related to individual, Accurate Prediction goes out criminal type and the crime probability of personnel to be measured, for public security cadres and police's emphasis investigation provide on-the-spot guidance.
Summary of the invention
Embodiment of the present invention technical problem to be solved is in that, it is provided that the method and system of a kind of suspicion of crime probabilistic forecasting, it is possible to Accurate Prediction goes out criminal type and the crime probability of personnel to be measured, provides on-the-spot guidance for public security cadres and police's emphasis investigation.
In order to solve above-mentioned technical problem, a kind of method embodiments providing suspicion of crime probabilistic forecasting, described method includes:
S1, obtain the relevant information of personnel to be measured, and the relevant information according to the described personnel to be measured got, the previous conviction type that described personnel to be measured are corresponding is determined from default information bank, and by each previous conviction type corresponding relevant historical data specifying index screening to go out respectively;
S2, determine the detection time, and according to the described each previous conviction type the determined corresponding relevant historical data specifying index screening to go out and detection time respectively, specify index all to carry out quantizing assignment to each in each previous conviction type;
S3, characteristic vector is set is formed by indexs whole in described appointment index, and according to each assignment specifying index corresponding in the characteristic vector of described formation and each previous conviction type, obtain the training sample database of each previous conviction type;
S4, the categorical attribute of the training sample database of the described each previous conviction type obtained all is divided into 1 and 0, and class probability when categorical attribute is 1 in each previous conviction type correspondence training sample database is respectively adopted logistic regression model is fitted, it is thus achieved that the crime probability of the corresponding each previous conviction type of described personnel to be measured; Wherein, described categorical attribute represents crime when being 1, described categorical attribute represents not crime when being 0.
Wherein, " relevant informations of personnel to be measured " in described step S1 specifically include personnel to be measured schooling, pursue an occupation, nationality, the age, marital status and height.
Wherein, described step S2 specifically includes:
Specify in each previous conviction type index be previous conviction number of times carry out quantizing assignment time, calculate in each previous conviction type each previous conviction number of times apart from the interval duration of described detection time, and according to each previous conviction number of times in the described each previous conviction type calculated apart from the interval duration of described detection time, it is thus achieved that the initial assignment that in each previous conviction type, each previous conviction number of times is corresponding respectively;
When described previous conviction number of times is 1, using the initial assignment corresponding when being 1 for previous conviction number of times in each previous conviction type of the described acquisition final assignment as its corresponding previous conviction number of times;
When described previous conviction number of times is more than 1, the initial assignment that each previous conviction number of times obtained described in same previous conviction type is corresponding respectively is added up, and the initial assignment of each previous conviction number of times correspondence respectively carries out the value after adding up according to same previous conviction type, obtain the final assignment of each previous conviction type corresponding previous conviction number of times respectively.
Wherein, described step S2 specifically also includes:
Specify in each previous conviction type index be online of high-risk period number of times carry out quantizing assignment time, determine the concrete period of described detection time, and count the ratio of each previous conviction type corresponding offender's online respectively under the described concrete period and the ratio of normal non-criminal online;
The ratio surfed the Net according to the corresponding offender of the described each previous conviction type counted difference and the ratio of normal non-criminal online, obtain the final assignment that in each previous conviction type, online of high-risk period number of times is corresponding.
Wherein, described step S2 specifically also includes:
Specify in each previous conviction type index be move in the high-risk period number of times carry out quantizing assignment time, determine the concrete period of described detection time, and count each previous conviction type ratio that corresponding offender moves in respectively and the ratio that normal non-criminal is moved under the described concrete period;
Corresponding offender moves in respectively according to the described each previous conviction type counted ratio and the ratio that normal non-criminal is moved in, obtaining the high-risk period in each previous conviction type moves in the final assignment that number of times is corresponding.
Wherein, described method farther includes:
On the crime probability of the corresponding each previous conviction type of personnel to be measured of described acquisition, correspondence gives weighter factor respectively, and calculate and give the crime probability of each previous conviction type after weighter factor and carry out the value added up, and further using the described accumulated value the calculated comprehensive crime index as described personnel to be measured.
The embodiment of the present invention additionally provides the system of a kind of suspicion of crime probabilistic forecasting, and described system includes:
Historical data acquiring unit, for obtaining the relevant information of personnel to be measured, and the relevant information according to the described personnel to be measured got, the previous conviction type that described personnel to be measured are corresponding is determined from default information bank, and by each previous conviction type corresponding relevant historical data specifying index screening to go out respectively;
Historical data assignment unit, is used for determining the detection time, and according to the described each previous conviction type the determined corresponding relevant historical data specifying index screening to go out and detection time respectively, specifies index all to carry out quantizing assignment to each in each previous conviction type;
Training sample database forms unit, is used for arranging characteristic vector and is formed by indexs whole in described appointment index, and according to the assignment that in the characteristic vector of described formation and each previous conviction type, respectively appointment index is corresponding, obtains the training sample database of each previous conviction type;
Crime probability prediction unit, for the categorical attribute of the training sample database of the described each previous conviction type obtained all is divided into 1 and 0, and class probability when categorical attribute is 1 in each previous conviction type correspondence training sample database is respectively adopted logistic regression model is fitted, it is thus achieved that the crime probability of the corresponding each previous conviction type of described personnel to be measured; Wherein, described categorical attribute represents crime when being 1, described categorical attribute represents not crime when being 0.
Wherein, described historical data assignment unit includes:
Previous conviction number of times assignment module, for specify in each previous conviction type index be previous conviction number of times carry out quantizing assignment time, calculate in each previous conviction type each previous conviction number of times apart from the interval duration of described detection time, and according to each previous conviction number of times in the described each previous conviction type calculated apart from the interval duration of described detection time, it is thus achieved that the initial assignment that in each previous conviction type, each previous conviction number of times is corresponding respectively;
When described previous conviction number of times is 1, using the initial assignment corresponding when being 1 for previous conviction number of times in each previous conviction type of the described acquisition final assignment as its corresponding previous conviction number of times;
When described previous conviction number of times is more than 1, the initial assignment that each previous conviction number of times obtained described in same previous conviction type is corresponding respectively is added up, and the initial assignment of each previous conviction number of times correspondence respectively carries out the value after adding up according to same previous conviction type, obtain the final assignment of each previous conviction type corresponding previous conviction number of times respectively.
Wherein, described historical data assignment unit also includes:
Online of high-risk period number of times assignment module, for specify in each previous conviction type index be online of high-risk period number of times carry out quantizing assignment time, determine the concrete period of described detection time, and count the ratio of each previous conviction type corresponding offender's online respectively under the described concrete period and the ratio of normal non-criminal online;
The ratio surfed the Net according to the corresponding offender of the described each previous conviction type counted difference and the ratio of normal non-criminal online, obtain the final assignment that in each previous conviction type, online of high-risk period number of times is corresponding.
Wherein, described historical data assignment unit also includes:
The high-risk period moves in number of times assignment module, move in the high-risk period for specifying in each previous conviction type index to be number of times carry out quantizing assignment time, determine the concrete period of described detection time, and count each previous conviction type ratio that corresponding offender moves in respectively and the ratio that normal non-criminal is moved under the described concrete period;
Corresponding offender moves in respectively according to the described each previous conviction type counted ratio and the ratio that normal non-criminal is moved in, obtaining the high-risk period in each previous conviction type moves in the final assignment that number of times is corresponding.
Implement the embodiment of the present invention, have the advantages that
In embodiments of the present invention, due to by the related information retrieval of the personnel to be measured historical data to its each previous conviction type corresponding, and the concrete period according to input current sensing time, appointment index from each previous conviction type is (such as previous conviction number of times, online number of times of high-risk period and high-risk period move in number of times etc.) quantize model analysis to corresponding historical data, obtain personnel to be measured and correspond to each previous conviction type crime probability in this concrete period, therefore criminal type and the crime probability predicted have accuracy, on-the-spot guidance can be provided for public security cadres and police's emphasis investigation.
Accompanying drawing explanation
In order to be illustrated more clearly that the embodiment of the present invention or technical scheme of the prior art, the accompanying drawing used required in embodiment or description of the prior art will be briefly described below, apparently, accompanying drawing in the following describes is only some embodiments of the present invention, for those of ordinary skill in the art, under the premise not paying creative work, the accompanying drawing obtaining other according to these accompanying drawings still falls within scope of the invention.
The flow chart of the method for a kind of suspicion of crime probabilistic forecasting that Fig. 1 provides for the embodiment of the present invention;
The structural representation of the system of a kind of suspicion of crime probabilistic forecasting that Fig. 2 provides for the embodiment of the present invention.
Detailed description of the invention
For making the object, technical solutions and advantages of the present invention clearly, below in conjunction with accompanying drawing, the present invention is described in further detail.
As it is shown in figure 1, the method for a kind of suspicion of crime probabilistic forecasting provided for the embodiment of the present invention, described method includes:
Step S1, obtain the relevant information of personnel to be measured, and the relevant information according to the described personnel to be measured got, the previous conviction type that described personnel to be measured are corresponding is determined from default information bank, and by each previous conviction type corresponding relevant historical data specifying index screening to go out respectively;
Detailed process is, in the emphasis troubleshooting procedure of public security cadres and police, the related data of on-the-spot inquiry personnel to be measured can be inputted by the identity card of hand-held identification equipment identification personnel to be measured or other input equipment (such as notebook), default information bank retrieves in (the police service information system such as backstage) relevant information of personnel to be measured, the relevant information of these personnel to be measured include but not limited to personnel to be measured schooling, pursue an occupation, nationality, the age, marital status and height. It should be noted that, during a certain loss of learning in the relevant information of personnel to be measured, can carry out perfect at the scene, such as nationality, marital status, height etc.
Simultaneously, also by default information bank, retrieve the previous conviction type that personnel to be measured are corresponding, and by each previous conviction type corresponding relevant historical data specifying index screening to go out respectively, this appointment index includes but not limited to that previous conviction number of times, online number of times of high-risk period and high-risk period move in number of times etc. It should be noted that, in the emphasis troubleshooting procedure of public security cadres and police, if not showing any previous conviction type, then show that personnel to be measured do not have previous conviction.
Step S2, determine the detection time, and according to the described each previous conviction type the determined corresponding relevant historical data specifying index screening to go out and detection time respectively, specify index all to carry out quantizing assignment to each in each previous conviction type;
Detailed process is, can be continuous data owing to specifying the data under index, being alternatively discrete data, it is also different that the historical data that therefore basis retrieves carries out the assignment procedure that quantizes of each appointment index, it is therefore desirable to continuous data and discrete data are processed by two ways respectively.
In one embodiment, it is intended that index includes the previous conviction number of times of continuous data class, and the online number of times of high-risk period of discrete data class and high-risk period move in number of times, implement step as follows:
(a) specify in each previous conviction type index be previous conviction number of times carry out quantizing assignment time, calculate the interval duration of each previous conviction number of times distance detection time in each previous conviction type, and according to the interval duration of each previous conviction number of times distance detection time in each previous conviction type calculated, it is thus achieved that the initial assignment that in each previous conviction type, each previous conviction number of times is corresponding respectively;
When previous conviction number of times is 1, using the initial assignment corresponding when being 1 for previous conviction number of times in each previous conviction type of the described acquisition final assignment as its corresponding previous conviction number of times;
When previous conviction number of times is more than 1, the initial assignment that each previous conviction number of times obtained described in same previous conviction type is corresponding respectively is added up, and carry out the value after adding up according to the initial assignment that previous conviction number of times each in same previous conviction type is corresponding respectively, obtain the final assignment of each previous conviction type corresponding previous conviction number of times respectively.
As an example, previous conviction number of times has time response, and the track occurred in the recent period more has reference significance, therefore adopts the continuous function that formula (1) is stated as time conversion function:
In formula (1), t0For the detection time; T is the time that each previous conviction number of times occurs; t0-t is interval duration, and wherein, interval duration takes double figures decimal, and its integer-bit is year, and after arithmetic point, first is day for second after the moon, arithmetic point.
As object stole the time of record of previous crime before 1 year, then this previous conviction track record can assignment be finally: ρ=e-1=0.37.
When having a plurality of track record in the historical data of previous conviction number of times, the method repeatedly accumulated can be adopted to embody its importance, namely adopt formula (2) to carry out accumulation calculating:
In formula (2), kmFor the amplification coefficient that expert sets, m is previous conviction number of times;
As: object has twice theft previous conviction, and the record time, then the final assignment of its record of previous crime was: 1.5 × (e before 1 year and before 4 years 3 months-1+e-4.25)=0.58.
(b) specify in each previous conviction type index be online of high-risk period number of times carry out quantizing assignment time, determine the concrete period of detection time, and count the ratio of each previous conviction type corresponding offender's online respectively under the concrete period and the ratio of normal non-criminal online;
The ratio surfed the Net according to the corresponding offender of each previous conviction type difference counted and the ratio of normal non-criminal online, obtain the final assignment that in each previous conviction type, online of high-risk period number of times is corresponding.
As an example, being in the 10:00-12:00 period in the concrete period, the historical statistical data 20% of swindle offender's online, non-offender is 5%, and this period is the swindle criminal high-risk online period. The assignment of online of high-risk period number of times adopts equation below (3) approximate representation:
In formula (3), k1T () is this swindle type ratio of offender's online, k within this period2The ratio that t people that () is the normal non-crime within this period of this swindle type surfs the Net.
As: swindle type in the final assignment of online of the high-risk period number of times that the concrete period is 10:00-12:00 is: μ=e(0.2-0.05)=1.16.
(c) specify in each previous conviction type index be move in the high-risk period number of times carry out quantizing assignment time, determine the concrete period of detection time, and count each previous conviction type ratio that corresponding offender moves in respectively and the ratio that normal non-criminal is moved under the concrete period;
Corresponding offender moves in respectively according to each previous conviction type counted ratio and the ratio that normal non-criminal is moved in, obtaining the high-risk period in each previous conviction type moves in the final assignment that number of times is corresponding.
As an example, being in the 23:00-24:00 period in the concrete period, the historical statistical data 20% that theft crime molecule is moved in, non-offender is 5%, and this period is theft, and criminal is high-risk moves in the period. The high-risk period moves in the assignment of number of times and can also adopt above-mentioned formula (3) approximate representation, is not repeating at this.
Step S3, characteristic vector is set is formed by indexs whole in described appointment index, and according to each assignment specifying index corresponding in the characteristic vector of described formation and each previous conviction type, obtain the training sample database of each previous conviction type;
Detailed process is, arranges eigen vector X={X1、X2、......、Xm, wherein, X1、X2、......、XmSpecify index for corresponding all of m respectively; And according to assignment specifying index each in the step S2 each previous conviction type obtained, count m the final assignment specifying index in same previous conviction type, thus obtaining the training sample database of each previous conviction type.
Step S4, the categorical attribute of the training sample database of the described each previous conviction type obtained all is divided into 1 and 0, and class probability when categorical attribute is 1 in each previous conviction type correspondence training sample database is respectively adopted logistic regression model is fitted, it is thus achieved that the crime probability of the corresponding each previous conviction type of described personnel to be measured; Wherein, described categorical attribute represents crime when being 1, described categorical attribute represents not crime when being 0.
Detailed process is, is all divided into 1 and 0 in each of the front in the categorical attribute of the training sample database of section's type, and this categorical attribute represents with dependent variable Y, and it is two classified variables, and value represents crime when being Y=1, and value is that Y=0 represents not crime.
M the independent variable affecting dependent variable Y value respectively respectively specifies index X1To Xm, the class probability that positive findings (i.e. crime) occurs under m independent variable (i.e. exposure factors) acts on adopts formula (4) to represent:
P=P (Y=1 | X1, X2..., Xm)(4)
By adopting Logistic conversion, make logit (P)=ln [P/ (1-P)], then Logistic regression model adopts formula (5) to represent::
Logit (P)=β0+β1X1+β2X2+…+βmXm(5)
By formula (5) after mathematic(al) manipulation, logistic regression model can pass through formula (6) and represent:
In formula (6), β0For constant term, β1, β2..., βmFor partial regression coefficient.
By gathering a number of training sample, above-mentioned formula (6) is trained, β can be obtained0, β1, β2..., βmValue Deng coefficient.
Due to β0, β1, β2..., βmCan matching obtaining Deng the value of coefficient, therefore, class probability when categorical attribute is 1 in each previous conviction type correspondence training sample database is respectively adopted logistic regression model when being fitted, and can directly obtain the crime probability of the corresponding each previous conviction type of personnel to be measured.
By the crime probabilistic forecasting to multiple criminal types, according to the needs analyzed of handling a case, simple weighted average can being adopted to obtain a comprehensive suspicion of crime desired value, therefore described method farther includes:
On the crime probability of the corresponding each previous conviction type of personnel to be measured of described acquisition, correspondence gives weighter factor respectively, and calculate and give the crime probability of each previous conviction type after weighter factor and carry out the value added up, and further using the described accumulated value the calculated comprehensive crime index as described personnel to be measured.
As in figure 2 it is shown, be in the embodiment of the present invention, it is provided that the system of a kind of suspicion of crime probabilistic forecasting, described system includes:
Historical data acquiring unit 210, for obtaining the relevant information of personnel to be measured, and the relevant information according to the described personnel to be measured got, the previous conviction type that described personnel to be measured are corresponding is determined from default information bank, and by each previous conviction type corresponding relevant historical data specifying index screening to go out respectively;
Historical data assignment unit 220, is used for determining the detection time, and according to the described each previous conviction type the determined corresponding relevant historical data specifying index screening to go out and detection time respectively, specifies index all to carry out quantizing assignment to each in each previous conviction type;
Training sample database forms unit 230, formed by indexs whole in described appointment index for arranging characteristic vector, and according to assignment specifying index corresponding each in the characteristic vector of described formation and each previous conviction type, obtain the training sample database of each previous conviction type;
Crime probability prediction unit 240, for the categorical attribute of the training sample database of the described each previous conviction type obtained all is divided into 1 and 0, and class probability when categorical attribute is 1 in each previous conviction type correspondence training sample database is respectively adopted logistic regression model is fitted, it is thus achieved that the crime probability of the corresponding each previous conviction type of described personnel to be measured; Wherein, described categorical attribute represents crime when being 1, described categorical attribute represents not crime when being 0.
Wherein, described historical data assignment unit 220 includes:
Previous conviction number of times assignment module 2201, for specify in each previous conviction type index be previous conviction number of times carry out quantizing assignment time, calculate in each previous conviction type each previous conviction number of times apart from the interval duration of described detection time, and according to each previous conviction number of times in the described each previous conviction type calculated apart from the interval duration of described detection time, it is thus achieved that the initial assignment that in each previous conviction type, each previous conviction number of times is corresponding respectively;
When described previous conviction number of times is 1, using the initial assignment corresponding when being 1 for previous conviction number of times in each previous conviction type of the described acquisition final assignment as its corresponding previous conviction number of times;
When described previous conviction number of times is more than 1, the initial assignment that each previous conviction number of times obtained described in same previous conviction type is corresponding respectively is added up, and the initial assignment of each previous conviction number of times correspondence respectively carries out the value after adding up according to same previous conviction type, obtain the final assignment of each previous conviction type corresponding previous conviction number of times respectively.
Wherein, described historical data assignment unit 220 also includes:
Online of high-risk period number of times assignment module 2202, for specify in each previous conviction type index be online of high-risk period number of times carry out quantizing assignment time, determine the concrete period of described detection time, and count the ratio of each previous conviction type corresponding offender's online respectively under the described concrete period and the ratio of normal non-criminal online;
The ratio surfed the Net according to the corresponding offender of the described each previous conviction type counted difference and the ratio of normal non-criminal online, obtain the final assignment that in each previous conviction type, online of high-risk period number of times is corresponding.
Wherein, described historical data assignment unit 220 also includes:
The high-risk period moves in number of times assignment module 2203, move in the high-risk period for specifying in each previous conviction type index to be number of times carry out quantizing assignment time, determine the concrete period of described detection time, and count each previous conviction type ratio that corresponding offender moves in respectively and the ratio that normal non-criminal is moved under the described concrete period;
Corresponding offender moves in respectively according to the described each previous conviction type counted ratio and the ratio that normal non-criminal is moved in, obtaining the high-risk period in each previous conviction type moves in the final assignment that number of times is corresponding.
Implement the embodiment of the present invention, have the advantages that
In embodiments of the present invention, due to by the related information retrieval of the personnel to be measured historical data to its each previous conviction type corresponding, and the concrete period according to input current sensing time, appointment index from each previous conviction type is (such as previous conviction number of times, online number of times of high-risk period and high-risk period move in number of times) quantize model analysis to corresponding historical data, obtain personnel to be measured and correspond to each previous conviction type crime probability in this concrete period, therefore criminal type and the crime probability predicted have accuracy, on-the-spot guidance can be provided for public security cadres and police's emphasis investigation.
It should be noted that in said system embodiment, each included system unit is carry out dividing according to function logic, but is not limited to above-mentioned division, as long as being capable of corresponding function; It addition, the concrete title of each functional unit is also only to facilitate mutually distinguish, it is not limited to protection scope of the present invention.
One of ordinary skill in the art will appreciate that all or part of step realizing in above-described embodiment method can be by the hardware that program carrys out instruction relevant and completes, described program can be stored in a computer read/write memory medium, described storage medium, such as ROM/RAM, disk, CD etc.
Above disclosed it is only present pre-ferred embodiments, certainly can not limit the interest field of the present invention, the equivalent variations therefore made according to the claims in the present invention with this, still belong to the scope that the present invention contains.
Claims (10)
1. the method for a suspicion of crime probabilistic forecasting, it is characterised in that described method includes:
S1, obtain the relevant information of personnel to be measured, and the relevant information according to the described personnel to be measured got, the previous conviction type that described personnel to be measured are corresponding is determined from default information bank, and by each previous conviction type corresponding relevant historical data specifying index screening to go out respectively;
S2, determine the detection time, and according to the described each previous conviction type the determined corresponding relevant historical data specifying index screening to go out and detection time respectively, specify index all to carry out quantizing assignment to each in each previous conviction type;
S3, characteristic vector is set is formed by indexs whole in described appointment index, and according to each assignment specifying index corresponding in the characteristic vector of described formation and each previous conviction type, obtain the training sample database of each previous conviction type;
S4, the categorical attribute of the training sample database of the described each previous conviction type obtained all is divided into 1 and 0, and class probability when categorical attribute is 1 in each previous conviction type correspondence training sample database is respectively adopted logistic regression model is fitted, it is thus achieved that the crime probability of the corresponding each previous conviction type of described personnel to be measured; Wherein, described categorical attribute represents crime when being 1, described categorical attribute represents not crime when being 0.
2. the method for claim 1, it is characterised in that " relevant informations of personnel to be measured " in described step S1 specifically include personnel to be measured schooling, pursue an occupation, nationality, the age, marital status and height.
3. the method for claim 1, it is characterised in that described step S2 specifically includes:
Specify in each previous conviction type index be previous conviction number of times carry out quantizing assignment time, calculate in each previous conviction type each previous conviction number of times apart from the interval duration of described detection time, and according to each previous conviction number of times in the described each previous conviction type calculated apart from the interval duration of described detection time, it is thus achieved that the initial assignment that in each previous conviction type, each previous conviction number of times is corresponding respectively;
When described previous conviction number of times is 1, using the initial assignment corresponding when being 1 for previous conviction number of times in each previous conviction type of the described acquisition final assignment as its corresponding previous conviction number of times;
When described previous conviction number of times is more than 1, the initial assignment that each previous conviction number of times obtained described in same previous conviction type is corresponding respectively is added up, and the initial assignment of each previous conviction number of times correspondence respectively carries out the value after adding up according to same previous conviction type, obtain the final assignment of each previous conviction type corresponding previous conviction number of times respectively.
4. the method for claim 1, it is characterised in that described step S2 specifically also includes:
Specify in each previous conviction type index be online of high-risk period number of times carry out quantizing assignment time, determine the concrete period of described detection time, and count the ratio of each previous conviction type corresponding offender's online respectively under the described concrete period and the ratio of normal non-criminal online;
The ratio surfed the Net according to the corresponding offender of the described each previous conviction type counted difference and the ratio of normal non-criminal online, obtain the final assignment that in each previous conviction type, online of high-risk period number of times is corresponding.
5. the method for claim 1, it is characterised in that described step S2 specifically also includes:
Specify in each previous conviction type index be move in the high-risk period number of times carry out quantizing assignment time, determine the concrete period of described detection time, and count each previous conviction type ratio that corresponding offender moves in respectively and the ratio that normal non-criminal is moved under the described concrete period;
Corresponding offender moves in respectively according to the described each previous conviction type counted ratio and the ratio that normal non-criminal is moved in, obtaining the high-risk period in each previous conviction type moves in the final assignment that number of times is corresponding.
6. the method for claim 1, it is characterised in that described method farther includes:
On the crime probability of the corresponding each previous conviction type of personnel to be measured of described acquisition, correspondence gives weighter factor respectively, and calculate and give the crime probability of each previous conviction type after weighter factor and carry out the value added up, and further using the described accumulated value the calculated comprehensive crime index as described personnel to be measured.
7. the system of a suspicion of crime probabilistic forecasting, it is characterised in that described system includes:
Historical data acquiring unit, for obtaining the relevant information of personnel to be measured, and the relevant information according to the described personnel to be measured got, the previous conviction type that described personnel to be measured are corresponding is determined from default information bank, and by each previous conviction type corresponding relevant historical data specifying index screening to go out respectively;
Historical data assignment unit, is used for determining the detection time, and according to the described each previous conviction type the determined corresponding relevant historical data specifying index screening to go out and detection time respectively, specifies index all to carry out quantizing assignment to each in each previous conviction type;
Training sample database forms unit, is used for arranging characteristic vector and is formed by indexs whole in described appointment index, and according to the assignment that in the characteristic vector of described formation and each previous conviction type, respectively appointment index is corresponding, obtains the training sample database of each previous conviction type;
Crime probability prediction unit, for the categorical attribute of the training sample database of the described each previous conviction type obtained all is divided into 1 and 0, and class probability when categorical attribute is 1 in each previous conviction type correspondence training sample database is respectively adopted logistic regression model is fitted, it is thus achieved that the crime probability of the corresponding each previous conviction type of described personnel to be measured; Wherein, described categorical attribute represents crime when being 1, described categorical attribute represents not crime when being 0.
8. the system as claimed in claim 1, it is characterised in that described historical data assignment unit includes:
Previous conviction number of times assignment module, for specify in each previous conviction type index be previous conviction number of times carry out quantizing assignment time, calculate in each previous conviction type each previous conviction number of times apart from the interval duration of described detection time, and according to each previous conviction number of times in the described each previous conviction type calculated apart from the interval duration of described detection time, it is thus achieved that the initial assignment that in each previous conviction type, each previous conviction number of times is corresponding respectively;
When described previous conviction number of times is 1, using the initial assignment corresponding when being 1 for previous conviction number of times in each previous conviction type of the described acquisition final assignment as its corresponding previous conviction number of times;
When described previous conviction number of times is more than 1, the initial assignment that each previous conviction number of times obtained described in same previous conviction type is corresponding respectively is added up, and the initial assignment of each previous conviction number of times correspondence respectively carries out the value after adding up according to same previous conviction type, obtain the final assignment of each previous conviction type corresponding previous conviction number of times respectively.
9. the system as claimed in claim 1, it is characterised in that described historical data assignment unit also includes:
Online of high-risk period number of times assignment module, for specify in each previous conviction type index be online of high-risk period number of times carry out quantizing assignment time, determine the concrete period of described detection time, and count the ratio of each previous conviction type corresponding offender's online respectively under the described concrete period and the ratio of normal non-criminal online;
The ratio surfed the Net according to the corresponding offender of the described each previous conviction type counted difference and the ratio of normal non-criminal online, obtain the final assignment that in each previous conviction type, online of high-risk period number of times is corresponding.
10. the system as claimed in claim 1, it is characterised in that described historical data assignment unit also includes:
The high-risk period moves in number of times assignment module, move in the high-risk period for specifying in each previous conviction type index to be number of times carry out quantizing assignment time, determine the concrete period of described detection time, and count each previous conviction type ratio that corresponding offender moves in respectively and the ratio that normal non-criminal is moved under the described concrete period;
Corresponding offender moves in respectively according to the described each previous conviction type counted ratio and the ratio that normal non-criminal is moved in, obtaining the high-risk period in each previous conviction type moves in the final assignment that number of times is corresponding.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610084782.0A CN105678428A (en) | 2016-01-28 | 2016-01-28 | Criminal suspicion probability prediction method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610084782.0A CN105678428A (en) | 2016-01-28 | 2016-01-28 | Criminal suspicion probability prediction method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105678428A true CN105678428A (en) | 2016-06-15 |
Family
ID=56304271
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610084782.0A Pending CN105678428A (en) | 2016-01-28 | 2016-01-28 | Criminal suspicion probability prediction method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105678428A (en) |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106952208A (en) * | 2017-03-17 | 2017-07-14 | 讯飞智元信息科技有限公司 | Crime automatic prediction method and system |
| CN107481177A (en) * | 2017-08-17 | 2017-12-15 | 航天科工智慧产业发展有限公司 | A kind of quick method and system realized a suspect and verify disposal guidance |
| CN107506936A (en) * | 2017-08-31 | 2017-12-22 | 苏州凌旭信息科技有限公司 | A kind of preliminary investigation standardised assessment system and method based on to suspect |
| CN108241910A (en) * | 2018-01-25 | 2018-07-03 | 广东惠禾科技发展有限公司 | A kind of Crime Spreading complex network node analysis method, apparatus and user terminal |
| CN108460504A (en) * | 2017-02-17 | 2018-08-28 | 张月妹 | A kind of supermarket's safe condition real-time visual evaluation and test algorithm |
| CN108596386A (en) * | 2018-04-20 | 2018-09-28 | 上海市司法局 | A kind of prediction convict repeats the method and system of crime probability |
| CN109063887A (en) * | 2018-06-13 | 2018-12-21 | 中国华戎科技集团有限公司 | A kind of crime hotspot prediction method and system |
| CN109492821A (en) * | 2018-11-20 | 2019-03-19 | 北京荣之联科技股份有限公司 | A kind of stability maintenance method for early warning and system, electronic equipment |
| CN109784525A (en) * | 2018-11-13 | 2019-05-21 | 北京码牛科技有限公司 | Method for early warning and device based on day vacant lot integration data |
| CN110009022A (en) * | 2019-03-26 | 2019-07-12 | 第四范式（北京）技术有限公司 | Prediction technique, device and the calculating equipment of drug addict's information |
| CN110351523A (en) * | 2019-07-22 | 2019-10-18 | 常州机电职业技术学院 | A kind of building video monitoring system and video monitoring method of adjustment |
| CN110580490A (en) * | 2018-06-11 | 2019-12-17 | 杭州海康威视数字技术股份有限公司 | Method, device and equipment for determining personnel behavior probability |
| CN110942192A (en) * | 2019-11-25 | 2020-03-31 | 中国联合网络通信集团有限公司 | Crime probability determination method and device |
| CN111062732A (en) * | 2019-12-24 | 2020-04-24 | 中科金审(北京)科技有限公司 | Method, device, equipment and storage medium for inquiring fake bearing production address |
| CN111783668A (en) * | 2020-07-01 | 2020-10-16 | 杭州海康威视系统技术有限公司 | Method, device and equipment for determining association degree between personnel and event |
| CN115936431A (en) * | 2022-11-28 | 2023-04-07 | 四川大学华西医院 | Crime risk assessment method, crime risk assessment device, computer equipment and readable storage medium |
-
2016
- 2016-01-28 CN CN201610084782.0A patent/CN105678428A/en active Pending
Cited By (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108460504A (en) * | 2017-02-17 | 2018-08-28 | 张月妹 | A kind of supermarket's safe condition real-time visual evaluation and test algorithm |
| CN106952208A (en) * | 2017-03-17 | 2017-07-14 | 讯飞智元信息科技有限公司 | Crime automatic prediction method and system |
| CN107481177A (en) * | 2017-08-17 | 2017-12-15 | 航天科工智慧产业发展有限公司 | A kind of quick method and system realized a suspect and verify disposal guidance |
| CN107506936A (en) * | 2017-08-31 | 2017-12-22 | 苏州凌旭信息科技有限公司 | A kind of preliminary investigation standardised assessment system and method based on to suspect |
| CN108241910A (en) * | 2018-01-25 | 2018-07-03 | 广东惠禾科技发展有限公司 | A kind of Crime Spreading complex network node analysis method, apparatus and user terminal |
| CN108596386A (en) * | 2018-04-20 | 2018-09-28 | 上海市司法局 | A kind of prediction convict repeats the method and system of crime probability |
| CN110580490A (en) * | 2018-06-11 | 2019-12-17 | 杭州海康威视数字技术股份有限公司 | Method, device and equipment for determining personnel behavior probability |
| CN109063887A (en) * | 2018-06-13 | 2018-12-21 | 中国华戎科技集团有限公司 | A kind of crime hotspot prediction method and system |
| CN109063887B (en) * | 2018-06-13 | 2021-09-17 | 中国华戎科技集团有限公司 | Crime hotspot prediction method and system |
| CN109784525A (en) * | 2018-11-13 | 2019-05-21 | 北京码牛科技有限公司 | Method for early warning and device based on day vacant lot integration data |
| CN109492821A (en) * | 2018-11-20 | 2019-03-19 | 北京荣之联科技股份有限公司 | A kind of stability maintenance method for early warning and system, electronic equipment |
| CN110009022A (en) * | 2019-03-26 | 2019-07-12 | 第四范式（北京）技术有限公司 | Prediction technique, device and the calculating equipment of drug addict's information |
| CN110351523A (en) * | 2019-07-22 | 2019-10-18 | 常州机电职业技术学院 | A kind of building video monitoring system and video monitoring method of adjustment |
| CN110351523B (en) * | 2019-07-22 | 2020-10-16 | 常州机电职业技术学院 | Building video monitoring system and video monitoring adjustment method |
| CN110942192A (en) * | 2019-11-25 | 2020-03-31 | 中国联合网络通信集团有限公司 | Crime probability determination method and device |
| CN111062732A (en) * | 2019-12-24 | 2020-04-24 | 中科金审(北京)科技有限公司 | Method, device, equipment and storage medium for inquiring fake bearing production address |
| CN111783668A (en) * | 2020-07-01 | 2020-10-16 | 杭州海康威视系统技术有限公司 | Method, device and equipment for determining association degree between personnel and event |
| CN115936431A (en) * | 2022-11-28 | 2023-04-07 | 四川大学华西医院 | Crime risk assessment method, crime risk assessment device, computer equipment and readable storage medium |
| CN115936431B (en) * | 2022-11-28 | 2023-10-20 | 四川大学华西医院 | Re-crime risk assessment method, device, computer equipment and readable storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105678428A (en) | Criminal suspicion probability prediction method and system | |
| CN110520874B (en) | Time-based ensemble machine learning model | |
| Zhuang et al. | Crime hot spot forecasting: A recurrent model with spatial and temporal information | |
| Law et al. | Analyzing Hotspots of Crime Using a B ayesian Spatiotemporal Modeling Approach: A Case Study of Violent Crime in the G reater T oronto A rea | |
| CN110674858B (en) | Traffic public opinion detection method based on space-time correlation and big data mining | |
| CN109214863B (en) | Method for predicting urban house demand based on express delivery data | |
| CN110889092A (en) | Short-time large-scale activity peripheral track station passenger flow volume prediction method based on track transaction data | |
| Zhang et al. | Anomaly detection of periodic multivariate time series under high acquisition frequency scene in IoT | |
| Hornyák et al. | AdaBoost algorithm could lead to weak results for data with certain characteristics | |
| Du et al. | Analysis of university fitness center data uncovers interesting patterns, enables prediction | |
| EP2682926B1 (en) | Traffic delay detection by mining ticket validation transactions | |
| Shukla et al. | Criminal Combat: Crime Analysis and Prediction Using Machine Learning | |
| CN111145535B (en) | Travel time reliability distribution prediction method under complex scene | |
| Yadav et al. | Modified ARIMA model for improving certainty in spatio-temporal crime event prediction | |
| Marevac et al. | Decision-making AI for customer worthiness and viability | |
| Sangeetha et al. | Crime Rate Prediction and Prevention: Unleashing the Power of Deep Learning | |
| Liu et al. | Inventory Management of Automobile After-sales Parts Based on Data Mining | |
| CN115965137B (en) | Specific object relevance prediction method, system, terminal and storage medium | |
| Okawa | Spatio-temporal Event Prediction via Deep Point Processes | |
| Izudheen | Short-term passenger count prediction for metro stations using LSTM network | |
| Wang | SELF-EXCITING POINT PROCESS FOR MODELLING TERROR ATTACK DATA | |
| Madhu | Crime Rate Prediction Using Time Series Forecasting | |
| Escobedo et al. | Comparing Regression Models to Predict Property Crime in High-Risk Lima Districts. | |
| Pretnar Žagar et al. | Model Evaluation: How to Accurately Evaluate Predictive Models | |
| Mason | Analysis of Virginia Crime Data of the Year 2016 Using Data Mining Techniques |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20160615 |