Embodiment
The assessment quantified to an object, is a problem being directed in many application scenarios.
, can be according to many factors of driver, such as illegal number of times, driving age for example, in driver's credit investigation system
Deng to driver progress Quantitative marking, to assess the fulfilling ability of the driver, if be adapted to drive certain vehicle, and
The driver can be educated or be discharged according to assessment result.
By taking driver's credit investigation system as an example, consider according to this many factors of driver can include it is as shown in table 1 below
Parameter, but following several parameters are citing, are not limited thereto in actual implementation:
The driver parameter of table 1
In table 1 above, driver identification can be the ID for one driver of unique mark, and zone of action represents this
Driver often drives vehicle in Kweiyang, the southeast of Guizhou Province of which regional activity, such as the example of table 1, can also have it is more other
Region;If driver is in a certain regional activity, can by table 1 to should the position in region be designated as " 1 ", otherwise can be designated as
" 0 " (a kind of exemplary mark mode).Illicit content therein can include it is polytype illegal, such as, and drink-driving,
Drive unlicensed vehicle, do not stop by regulation, the illegal of each type can use a corresponding illegal code table
Show, can be " 1 " by the position mark in corresponding table 1 if driver triggers certain Criminal type.In table 1, one
The corresponding data line of individual driver is properly termed as the sample of the driver, and { S1,5,1,0,1,0 } as escribed above is to drive
The person of sailing a S1 sample data;Same driver can have multiple samples in table 1.
In the example of the disclosure, driver parameter can be divided into the following two kinds type:
Continuous variable:Continuous variable be it is a kind of can have continuous value variable, for example, " driving age " is one kind
Continuous variable, the value of driving age can include each values such as " 2,2.5,3,3.5,4 ... ".
Character type variable:Character type variable be it is a kind of can have discrete value variable, for example, with the activity in table 1
Exemplified by region, multiple centrifugal pumps such as " Kweiyang, the southeast of Guizhou Province " can be included by being somebody's turn to do the value of " zone of action ";" illicit content " can also
Including " multiple discrete Criminal types such as A, B ".
Fig. 1 illustrates the flow of the object selection method of disclosure example, by taking above-mentioned driver's credit investigation system as an example, leads to
Quantitative evaluation can be carried out to each driver by crossing this method, and satisfaction spy is selected in multiple drivers according to assessment result
The driver of fixed condition.In the present example, object is " driver ".
In a step 101, the image parameter of object is obtained, the image parameter includes:Character type variable and continuous type become
Amount.
In this example, image parameter is to be used for the foundation as quantitative evaluation object, whether character type variable or company
Ideotype variable, will all be used to quantitative evaluation object.Such as, in driver parameter, can by the driving age, zone of action, it is illegal in
Various types of variables such as appearance, all for subsequent step in the quantization of driver.In conventional process mode, be typically according to
Continuous variable is quantified, and causes the assessment to driver less accurate, and this example is by by continuous variable and character type
Variable all includes the quantization to driver, will make it that the assessment to driver is more accurate.
In a step 102, to the character type variable and continuous variable, carried out according to predetermined quantitative mode at quantization
Reason, obtains the corresponding quantized value of the object.
In this step, character type variable and continuous variable can be carried out quantifying, and the quantification manner of these variables
Quantification treatment can be carried out with identical according to predetermined quantification manner.So by various types of variable parameters of driver all
Participate in quantify, it is possible to so that each driver can carry out quantitative evaluation according to unified mode, it is ensured that each driver it
Between quantization compare more fair.
In step 103, if the quantized value reaches the number range of predetermined destination object, the object is selected to make
For the destination object.
This step can be according to the quantized value of object, and whether judge the object is required destination object;Can be advance
The number range of sets target object, if the quantized value that step 102 is obtained is within the range, it is target that can determine the object
Object.For example, in driver's credit investigation system, if purpose is to select the driver with potential safety hazard, will can have
The driver for having potential safety hazard is referred to as blacklist object, and sets the number range of blacklist object, and the number range can be
Quantized value L1~quantized value L2, wherein number range maximum L2, if the quantized value of a driver evaluation>WL2, herein
W value can then determine that the attribute of the driver is more biased towards in blacklist object close to 1, such as w=0.9.Need explanation
, after the quantized value of object is obtained, how can independently be set according to quantized value selection target object, this example is not limited
System, an above-mentioned simply citing.
The object selection method provided in disclosure example, by the way that continuous variable and character type variable are all included to driving
The quantization for the person of sailing, considering for this many factors, will cause the assessment more science to driver, so as to can also carry
The science of high Object Selection.
In another example, the character type variable in image parameter can also be refined, be divided into first kind word
Accord with variable and Equations of The Second Kind character variable.Wherein, the quantity for the types of variables that first kind character variable includes is less than predetermined quantity, and
Variable described in character type variable outside first kind character variable, is Equations of The Second Kind character variable.
For example, by taking " zone of action " in driver parameter as an example, the zone of action of general each provinces and cities is generally less than
20, if zone of action is referred to as into first kind character variable, including the region such as Kweiyang, the southeast of Guizhou Province be referred to as variable
Type, then the quantity for the types of variables that the first kind character variable includes is limited, and 20 are less than than described above.
In specific implementation, a predetermined quantity can be set for types of variables, if the number for the types of variables that a certain character variable includes
Amount is less than the predetermined quantity, then the character variable can be classified as to first kind character variable;Otherwise, if what character variable included
The quantity of types of variables is more than or equal to predetermined quantity, then the character variable can be classified as to Equations of The Second Kind character variable.Such as, character
Included Criminal type quantity is more in variable " illicit content ", can be used as Equations of The Second Kind character variable.
Though in addition, first kind character variable or Equations of The Second Kind character variable, can include it is multiple, as described above
" zone of action " is one of which first kind character variable, can also include other first kind character variables;Equally, it is above-mentioned
" illicit content " is one of which Equations of The Second Kind character variable, can also include its dependent variable.Also, first kind character variable is not only
Including types of variables quantity it is relatively fewer, the quantity of the variable-value under each types of variables is also typically larger than a fixed number
Amount, such as, the region such as Kweiyang, southeast of Guizhou Province belongs to types of variables, under each types of variables such as " Kweiyang ", may have and much drive
The sample for the person of sailing, sample size is more;And the quantity for the types of variables that Equations of The Second Kind character variable not only includes is relatively more, each
The quantity of variable-value under types of variables is also generally less, and the value under some typess of variables may be lacked, such as is disobeying
Driver's sample may be not present under method code B, i.e., no driver makes such mistake.
Fig. 2, which is illustrated in the handling process of another object selection method, this method, more to be refined according to above-mentioned
Parametric classification carries out quantitative evaluation to driver, and by taking the assessment to one of driver as an example, this method can include as follows
Following execution sequence is not limited in step, but specific implementation:
In step 201, respectively under each types of variables, the first ratio and the second ratio are obtained.
By taking one of driver as an example, the parameter of the driver includes the variable of many types, than portion as shown in table 1
Variation per minute;, can be by each variable in continuous variable, first kind character variable, Equations of The Second Kind character variable in this example
Type is referred to as one " level value ", such as in table 1, and " driving age " is a level value, and " Kweiyang " is a level value, " Guizhou Province
The southeast " is a level value, and " illegal code A " is a level value, and " illegal code B " is a level value.
In this step, the first ratio and the second ratio of the driver respectively under each level value, can be calculated.Its
In, the first ratio is ratio of the quantity with all bad total sample number amounts of the bad sample of driver, and the second ratio is driver
The ratio of the quantity of good sample and all good total sample number amounts.
By taking " Kweiyang " as an example, the driver assessed in this example may have multiple samples under " Kweiyang " level value
This, such as the S1 in table 1 just includes two samples, counts good sample size and bad sample number of the driver under " Kweiyang "
Amount, wherein, " good sample " is the sample that accident does not occur, and " bad sample " is the sample for the accident that there occurs;Also need to statistics " expensive
All good sample sizes and all bad sample size (i.e. Kweiyang under the good sample and bad sample of all users of the sun " under level value
This).
By the quantity of the bad sample of driver compared with all bad total sample number amounts, the first ratio is obtained.For example,Table
Show the number of the bad sample of the driver in some level value,For the sum of all bad samples in the corresponding level value.
Wherein m=1 is expressed as first kind character variable, and m=2 is Equations of The Second Kind character variable, and m=3 is continuous variable.
It is then the first ratio, each level value in first kind character variable, Equations of The Second Kind character variable, continuous variable can be according to
The formula is calculated.
By the quantity of the good sample of driver compared with all good total sample number amounts, the second ratio is obtained.For example,Table
Show the number of the good sample of the driver in some level value,For the sum of all good samples in the corresponding level value.
Wherein m=1 is expressed as first kind character variable, and m=2 is Equations of The Second Kind character variable, and m=3 is continuous variable.
It is then the second ratio, each level value in first kind character variable, Equations of The Second Kind character variable, continuous variable can be according to
The formula is calculated.
In step 202., according to first ratio and the second ratio, the son amount under each types of variables is respectively obtained
Change value.
For example, the sub- quantized value under one level value of correspondence can be calculated according to two ratios obtained in step 201.
A kind of exemplary mode can be:The numerical value that the formula is calculated is correspondence water
Level values i quantized value.
In step 203, the sub- quantized value summation of each types of variables under first kind character variable is obtained into the first character
Quantized value, obtains the second character quantized value, by even by the sub- quantized value summation of each types of variables under Equations of The Second Kind character variable
The sub- quantized value summation of each types of variables under ideotype variable obtains continuous type quantized value.
For example, multiple level values can be included under first kind character variable, the son calculated under each level value i is quantified
Value is added, and can obtain the first character quantized value, such as,It is that first kind character becomes
Measure the first character quantized value summed under multiple level values.Similarly, the second character quantized value and continuous type quantized value can be obtained.
In step 204, the first character quantized value, the second character quantized value and continuous type quantized value are summed, obtained
To the corresponding quantized value of the object.
This step can be by the first character quantized value obtained in above-mentioned steps 203, the second character quantized value and continuous type
Quantized value is summed, wherein, the respective weight of this three aspect factor can be set in summation, the weight of the first character quantized value can
With the weight higher than the second character quantized value.
For example, can be calculated according to equation below:
Wherein, scale is the corresponding quantized value of object, is also the quantization score value for the driver that this example is assessed.1-a
Can be the weight of the first character quantized value, a is the weight of the second character quantized value.Due to the level value of first kind character variable
The sample of less and each level value is more, can assign higher weight, and the level value of Equations of The Second Kind character variable is too many, than
1000 are such as usually more than, and can extremely be lacked in the sample under some level values, it is smaller that such variable is more suitable for imparting
Weight.In one example, α numerical value can be set to be less than 0.1, it is not too many variable power that can so cause level value
Weight is bigger.
In step 205, if the quantized value reaches the number range of predetermined destination object, the object is selected to make
For the destination object.
The object selection method provided in disclosure example, by the way that continuous variable and character type variable are all included to driving
The quantization for the person of sailing, considering for this many factors, will make it that the assessment to driver is more accurate, so as to can also carry
The science of high Object Selection;And by the way that character type variable is refined so that consider different type Variable Factors and exist
The not same-action in terms of object quantization is weighed, the science that object quantifies and selected is further increased.
With reference to above-mentioned embodiment of the method, the driving data that a kind of automobile driver is provided in disclosure example is handled
Method, the driving data for example can include above-described embodiment in refer to character type variable (for example, driving-activity region, disobey
Method content) and continuous variable (for example, driving age), driving data processing here can be become according to driving age and illicit content etc.
Amount is estimated to driver.This method for example can apply to driver's credit investigation system.
The driving data processing method of the automobile driver can include following processing, wherein, following process step is only
Briefly describe, detailed description can be combined referring to above method embodiment:
Respectively under each types of variables in driving data, the first ratio and the second ratio, first ratio are obtained
Value is ratio of the quantity with all bad total sample number amounts of the bad sample of the driver under the types of variables, and the second ratio is institute
State the ratio of the quantity and all good total sample number amounts of the good sample of driver under types of variables;
According to first ratio and the second ratio, the sub- quantized value under each types of variables is respectively obtained;
According to the sub- quantized value under each types of variables, the corresponding quantized value of driver is obtained.
The driving data processing method of this automobile driver, due to by the character type variable in driving data and continuously
Type variable all incorporates the quantitative evaluation to driver, therefore can more accurately evaluate driver.
In another example, character type variable can be refined, including:First kind character variable and Equations of The Second Kind character
Variable;The quantity for the types of variables that the first kind character variable includes is less than predetermined quantity;Described in the character type variable
Variable outside first kind character variable, is Equations of The Second Kind character variable;The first kind character variable, Equations of The Second Kind character variable and
Continuous variable, respectively including multiple typess of variables.
In the example, when according to the sub- quantized value under each types of variables, obtaining the corresponding quantized value of driver, specifically
It can include:The sub- quantized value summation of each types of variables under the first kind character variable is obtained into the first character quantized value,
The sub- quantized value summation of each types of variables under the Equations of The Second Kind character variable is obtained into the second character quantized value, by the company
The sub- quantized value summation of each types of variables under ideotype variable obtains continuous type quantized value.By the first character quantized value,
Second character quantized value and the summation of continuous type quantized value, obtain the corresponding quantized value of the driver.
Certainly, after the corresponding quantized value of driver is obtained, it can also further apply the quantized value, such as, according to
Quantized value selection target driver, if the quantized value reaches the number range of predetermined target driver, drives described in selection
The person of sailing is used as the target driver.In different application scenarios, preferable driver can be selected according to quantized value, also may be used
To select poor driver according to quantized value, and how a variety of setting means can also be had according to quantized value selection, do not done
Limitation.
A kind of object selecting device is additionally provided in order to realize in above-mentioned object selection method, the example of the disclosure, such as
Shown in Fig. 3, the object selecting device can include:Parameter acquisition module 31, quantification treatment module 32 and Object Selection module 33.
Parameter acquisition module 31, the image parameter for obtaining object, the image parameter includes:Character type variable and company
Ideotype variable;
Quantification treatment module 32, for the character type variable and continuous variable, entering according to predetermined quantitative mode
Row quantification treatment, obtains the corresponding quantized value of the object;
When Object Selection module 33, number range for reaching predetermined destination object in the quantized value, then select
The object is used as the destination object.
In one example, the character type variable, including:First kind character variable and Equations of The Second Kind character variable;It is described
The quantity for the types of variables that first kind character variable includes is less than predetermined quantity;First kind character described in the character type variable
Variable outside variable, is Equations of The Second Kind character variable.
In one example, as shown in figure 4, the first kind character variable, Equations of The Second Kind character variable and continuous variable,
Include multiple typess of variables respectively;The quantification treatment module 32, including:
Ratio determination sub-module 321, under each types of variables, obtaining the first ratio and the second ratio respectively,
First ratio is the quantity and the bad total sample number amount of all objects of the bad sample of the object under the types of variables
Ratio, second ratio is the quantity and the good total sample number of all objects of the good sample of the object under the types of variables
The ratio of amount;
Quantify determination sub-module 322, for according to first ratio and the second ratio, respectively obtaining each variable class
Sub- quantized value under type;
Summation process submodule 323, for the sub- quantized value of each types of variables under the first kind character variable to be asked
With obtain the first character quantized value, the summation of the sub- quantized value of each types of variables under the Equations of The Second Kind character variable is obtained the
Two character quantized values, continuous type quantized value is obtained by the sub- quantized value summation of each types of variables under the continuous variable;
Quantized result submodule 324, for the first character quantized value, the second character quantized value and continuous type to be quantified
Value summation, obtains the corresponding quantized value of the object.
In one example, quantized result submodule 324, for by the first character quantized value, the second character amount
When change value and the summation of continuous type quantized value, the weight of the first character quantized value is higher than the weight of the second character quantized value.
In one example, Object Selection module 33, if being less than predetermined threshold value, the threshold specifically for the quantized value
It is worth the number range maximum for blacklist object, then selects the object as blacklist object.
Device or module that above-described embodiment is illustrated, can specifically be realized by computer chip or entity, or by with
The product of certain function is realized.A kind of typically to realize that equipment is computer, the concrete form of computer can be personal meter
Calculation machine, laptop computer, cell phone, camera phone, smart phone, personal digital assistant, media player, navigation are set
It is any several in standby, E-mail receiver/send equipment, game console, tablet PC, wearable device or these equipment
The combination of equipment.
For convenience of description, it is divided into various modules during description apparatus above with function to describe respectively.Certainly, this is being implemented
The function of each module can be realized in same or multiple softwares and/or hardware when open.
The preferred embodiment of the disclosure is the foregoing is only, not to limit the disclosure, all essences in the disclosure
God is with principle, and any modification, equivalent substitution and improvements done etc. should be included within the scope of disclosure protection.