CN108153591A - Data flow real-time processing method, device and storage medium - Google Patents
Data flow real-time processing method, device and storage medium Download PDFInfo
- Publication number
- CN108153591A CN108153591A CN201711274270.1A CN201711274270A CN108153591A CN 108153591 A CN108153591 A CN 108153591A CN 201711274270 A CN201711274270 A CN 201711274270A CN 108153591 A CN108153591 A CN 108153591A
- Authority
- CN
- China
- Prior art keywords
- data
- value
- characteristic value
- feature value
- data flow
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5005—Allocation of resources, e.g. of the central processing unit [CPU] to service a request
- G06F9/5011—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resources being hardware resources other than CPUs, Servers and Terminals
- G06F9/5016—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resources being hardware resources other than CPUs, Servers and Terminals the resource being the memory
Landscapes
- Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a kind of data flow real-time processing method, device and storage medium, the method includes：Receive data flow in nth data, and from default storage unit obtain before the corresponding first history feature value of 1 data of n；Corresponding first current characteristic value of n data before being calculated according to the nth data and the corresponding first history feature value of 1 data of the preceding n；First current characteristic value is shown, first current characteristic value is replaced to the characteristic value stored in the default storage unit, the n is carried out certainly plus 1 calculates, and return to receiving step.Since total data need not be preserved, according only to current data and history feature value, the corresponding current characteristic value of total data can be calculated, can not only show current characteristic value in real time, and can largely reduce calculation amount and reduce fortune and deposit.
Description
Technical field
The present invention relates to a kind of Data Stream Processing field more particularly to data flow real-time processing method, device and storages to be situated between
Matter.
Background technology
With transition of the conventional appliances to light weight internet medical instrument, more and more algorithms are in the form of software
It has been transplanted in embedded mobile system.So, how to calculate, in the mobile equipment that storage resource is limited, realized with software
Such as the function as traditional large medical apparatus hardware, just become significant challenge.Moreover, at wider signal
System regions are managed, more and more equipment all develop toward lighter direction.Due to the lightweight of hardware, software will undertake biography
The system same function of hardware：Streaming data carries out duct type and handles in real time.
Current data flow processing method is that all continuous storage is deposited to fortune the data flow inputted in real time, every time fortune
All data are calculated in depositing, and could export desired value, this method drawback is it is clear that calculation amount is larger, with number
According to continually entering for stream, fortune is deposited can overflow eventually, it is difficult to accomplish that low fortune is deposited.It, may due to the limitation of property inside objectives value
Can not or be difficult to accomplish to export in real time, therefore, it is necessary to can take into account low fortune in a kind of Data Stream Processing to deposit reality with real-time
Existing method.
Invention content
It is a primary object of the present invention to provide a kind of data flow real-time processing method, device and storage medium, it is intended to solve
It is certainly difficult to take into account low fortune in Data Stream Processing in the prior art to deposit and the technical issues of real-time.
To achieve the above object, the present invention provides a kind of data flow real-time processing method, the described method comprises the following steps：
The nth data in data flow is received, and n-1 data corresponding first are gone through before acquisition from default storage unit
History characteristic value, the n are the integer not less than 1；
N data pair before being calculated according to the nth data and the corresponding first history feature value of the preceding n-1 data
The first current characteristic value answered；
First current characteristic value is shown, first current characteristic value is replaced into the default storage unit
The characteristic value of middle storage carries out the n from plus 1 calculates, and return to the nth data in the reception data flow, and from pre-
If before being obtained in storage unit the step of n-1 data corresponding first history feature value.
Preferably, it is described receive data flow in nth data, and from default storage unit obtain before n-1 data
Before corresponding first history feature value, the method further includes：
Data processing instructions input by user are received, target signature, the target are extracted from the data processing instructions
It is characterized as the corresponding feature of characteristic value to be shown；
The classification belonging to the target signature is obtained, when the target signature is the first category feature, performs the reception
Nth data in data flow, and from default storage unit obtain before the corresponding first history feature value of n-1 data step
Suddenly；
Correspondingly, it is described to show first current characteristic value, it specifically includes：
Using first current characteristic value as first kind characteristic value, and show the first kind characteristic value, described first
Category feature value is the corresponding characteristic value of first category feature.
Preferably, the classification obtained belonging to the target signature, after the target signature is the first category feature,
The method further includes：
Search corresponding with target signature target formula in mapping relations, in the mapping relations comprising feature and
The correspondence of formula；
Correspondingly, it is described to be calculated according to the nth data with the corresponding first history feature value of the preceding n-1 data
Corresponding first current characteristic value of preceding n data, specifically includes：
According to the target formula to the nth data and the corresponding first history feature value of the preceding n-1 data
It is calculated, obtains corresponding first current characteristic value of the preceding n data.
Preferably, after the classification obtained belonging to the target signature, the method further includes：
When the target signature is the second category feature, the nth data in the data flow is received, and preset from described
The corresponding second history feature value of n-1 data before being obtained in storage unit；
N data pair before being calculated according to the nth data and the corresponding second history feature value of the preceding n-1 data
The the second category feature value answered, the second category feature value are the corresponding characteristic value of second category feature；
The second category feature value is shown, the second history feature value is updated, the n is carried out certainly
Add 1 calculating, and return to the nth data received in the data flow, and n-1 number before being obtained from default storage unit
The step of according to corresponding second history feature value.
Preferably, it is described to be calculated according to the nth data with the corresponding second history feature value of the preceding n-1 data
The corresponding second category feature value of preceding n data, specifically includes：
The nth data and the corresponding second history feature value of the preceding n-1 data are calculated, obtain preceding n
Corresponding second current characteristic value of a data；
The preceding n data corresponding second are calculated according to corresponding second current characteristic value of the preceding n data
Category feature value.
Preferably, it is described that the second history feature value is updated, it specifically includes：
Second current characteristic value is replaced to the second history feature value stored in the default storage unit.
Preferably, it is described return it is described receive data flow in nth data, and from default storage unit obtain before n-
After the step of 1 data corresponding first history feature value, the method further includes：
The nth data is deleted.
Preferably, after the classification obtained belonging to the target signature, the method further includes：
When the target signature is third category feature, the nth data and (n+1)th number in the data flow are received
According to, and obtain from the default storage unit third history feature value of (n-1)th data and the nth data；
The third current characteristic value of the nth data and (n+1)th data is calculated, by the third history feature
Value is compared with the third current characteristic value, and third category feature value, the third category feature value are obtained according to comparison result
For the corresponding characteristic value of the third category feature；
The third category feature value is shown, the third current characteristic value is replaced in the default storage unit
The third history feature value of storage carries out the n from plus 1 calculates, and return receive nth data in the data flow and
The step of (n+1)th data.
In addition, to achieve the above object, the present invention also provides a kind of data flow real-time processing device, described device includes：
Memory, processor and the data flow real-time handler that can be run on the memory and on the processor is stored in,
The step of data flow real-time handler realizes the data flow real-time processing method when being performed by the processor.
In addition, to achieve the above object, the present invention also provides a kind of storage medium, data are stored on the storage medium
Real-time handler is flowed, the data flow real-time handler realizes the data flow real-time processing method when being executed by processor
The step of.
In the present invention, by receiving the nth data in data flow, and from default storage unit n-1 before acquisition
The corresponding first history feature value of data, the n are the integer not less than 1；According to the nth data and the preceding n-1
Corresponding first current characteristic value of n data before the corresponding first history feature value of data calculates；By first current signature
Value is shown, first current characteristic value is replaced to the characteristic value stored in the default storage unit, and the n is carried out
From adding 1 calculating, and the nth data in the reception data flow is returned to, and n-1 data before being obtained from default storage unit
The step of corresponding first history feature value.Since total data need not be preserved, according only to current data and history feature value, i.e.,
The corresponding current characteristic value of total data can be calculated, can not only show current characteristic value in real time, and can be very
Calculation amount is reduced in big degree and reduces fortune and is deposited.
Description of the drawings
Fig. 1 is the data flow real-time processing device structural representation for the hardware running environment that the embodiment of the present invention is related to
Figure；
Fig. 2 is the flow diagram of data flow real-time processing method first embodiment of the present invention；
Fig. 3 is the flow diagram of data flow real-time processing method second embodiment of the present invention；
Fig. 4 is the flow diagram of data flow real-time processing method 3rd embodiment of the present invention；
Fig. 5 is the flow diagram of data flow real-time processing method fourth embodiment of the present invention；
Fig. 6 is data flow real-time processing method noise reduction process flow diagram of the present invention.
The embodiments will be further described with reference to the accompanying drawings for the realization, the function and the advantages of the object of the present invention.
Specific embodiment
It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, it is not intended to limit the present invention.
The solution of the embodiment of the present invention is mainly：By receiving the nth data in data flow, and from default storage
The corresponding first history feature value of n-1 data before being obtained in unit, the n are the integer not less than 1；According to described n-th
Corresponding first current characteristic value of n data before data and the corresponding first history feature value of the preceding n-1 data calculate；It will
First current characteristic value is shown, first current characteristic value is replaced to the spy stored in the default storage unit
Value indicative carries out the n from plus 1 calculates, and return to the nth data in the reception data flow, and from default storage unit
Before middle acquisition the step of n-1 data corresponding first history feature value.Since total data need not be preserved, according only to current number
According to history feature value, the corresponding current characteristic value of total data can be calculated, can not only show current spy in real time
Value indicative, and can largely reduce calculation amount and reduce fortune and deposit.
With reference to Fig. 1, Fig. 1 is the data flow real-time processing device knot of hardware running environment that the embodiment of the present invention is related to
Structure schematic diagram.
As shown in Figure 1, the data flow real-time processing device can include：Processor 1001, such as CPU, communication bus
1002nd, user interface 1003, network interface 1004, memory 1005.Wherein, communication bus 1002 be used to implement these components it
Between connection communication.User interface 1003 can include display screen (Display), and optional user interface 1003 can also include mark
Wireline interface, the wireless interface of standard.Network interface 1004 can optionally include standard wireline interface and wireless interface (such as WI-
FI interfaces).Memory 1005 can be high-speed RAM memory or the memory (non-volatile of stabilization
), such as magnetic disk storage memory.Memory 1005 optionally can also be the storage service independently of aforementioned processor 1001
Device.
It will be understood by those skilled in the art that the structure shown in Fig. 1 is not formed handles dress in real time to the data flow
The restriction put can include either combining certain components or different components arrangement than illustrating more or fewer components.
As shown in Figure 1, as operating system, network communication mould can be included in a kind of memory 1005 of storage medium
Block, Subscriber Interface Module SIM and data flow real-time handler.
In the construction shown in fig. 1, network interface 1004 is mainly used for Connection Service device, and data are carried out with the server
Communication；User interface 1003 is mainly used for connecting terminal, with terminal into row data communication；Described device is adjusted by processor 1001
With the data flow real-time handler stored in memory 1005, and perform following operate：
The nth data in data flow is received, and n-1 data corresponding first are gone through before acquisition from default storage unit
History characteristic value, the n are the integer not less than 1；
N data pair before being calculated according to the nth data and the corresponding first history feature value of the preceding n-1 data
The first current characteristic value answered；
First current characteristic value is shown, first current characteristic value is replaced into the default storage unit
The characteristic value of middle storage carries out the n from plus 1 calculates, and return to the nth data in the reception data flow, and from pre-
If before being obtained in storage unit the step of n-1 data corresponding first history feature value.
Further, processor 1001 can call the data flow real-time handler stored in memory 1005, also hold
Row is following to be operated：
Data processing instructions input by user are received, target signature, the target are extracted from the data processing instructions
It is characterized as the corresponding feature of characteristic value to be shown；
The classification belonging to the target signature is obtained, when the target signature is the first category feature, performs the reception
Nth data in data flow, and from default storage unit obtain before the corresponding first history feature value of n-1 data step
Suddenly；
Correspondingly, it is described to show first current characteristic value, it specifically includes：
Using first current characteristic value as first kind characteristic value, and show the first kind characteristic value, described first
Category feature value is the corresponding characteristic value of first category feature.
Further, processor 1001 can call the data flow real-time handler stored in memory 1005, also hold
Row is following to be operated：
Search corresponding with target signature target formula in mapping relations, in the mapping relations comprising feature and
The correspondence of formula；
Correspondingly, it is described to be calculated according to the nth data with the corresponding first history feature value of the preceding n-1 data
Corresponding first current characteristic value of preceding n data, specifically includes：
According to the target formula to the nth data and the corresponding first history feature value of the preceding n-1 data
It is calculated, obtains corresponding first current characteristic value of the preceding n data.
Further, processor 1001 can call the data flow real-time handler stored in memory 1005, also hold
Row is following to be operated：
When the target signature is the second category feature, the nth data in the data flow is received, and preset from described
The corresponding second history feature value of n-1 data before being obtained in storage unit；
N data pair before being calculated according to the nth data and the corresponding second history feature value of the preceding n-1 data
The the second category feature value answered, the second category feature value are the corresponding characteristic value of second category feature；
The second category feature value is shown, the second history feature value is updated, the n is carried out certainly
Add 1 calculating, and return to the nth data received in the data flow, and n-1 number before being obtained from default storage unit
The step of according to corresponding second history feature value.
Further, processor 1001 can call the data flow real-time handler stored in memory 1005, also hold
Row is following to be operated：
The nth data and the corresponding second history feature value of the preceding n-1 data are calculated, obtain preceding n
Corresponding second current characteristic value of a data；
The preceding n data corresponding second are calculated according to corresponding second current characteristic value of the preceding n data
Category feature value.
Further, processor 1001 can call the data flow real-time handler stored in memory 1005, also hold
Row is following to be operated：
Second current characteristic value is replaced to the second history feature value stored in the default storage unit.
Further, processor 1001 can call the data flow real-time handler stored in memory 1005, also hold
Row is following to be operated：
The nth data is deleted.
Further, processor 1001 can call the data flow real-time handler stored in memory 1005, also hold
Row is following to be operated：
When the target signature is third category feature, the nth data and (n+1)th number in the data flow are received
According to, and obtain from the default storage unit third history feature value of (n-1)th data and the nth data；
The third current characteristic value of the nth data and (n+1)th data is calculated, by the third history feature
Value is compared with the third current characteristic value, and third category feature value, the third category feature value are obtained according to comparison result
For the corresponding characteristic value of the third category feature；
The third category feature value is shown, the third current characteristic value is replaced in the default storage unit
The third history feature value of storage carries out the n from plus 1 calculates, and return receive nth data in the data flow and
The step of (n+1)th data.
In the present embodiment, by receiving the nth data in data flow, and n-1 before being obtained from default storage unit
The corresponding first history feature value of a data, the n are the integer not less than 1；According to the nth data and the preceding n-1
Corresponding first current characteristic value of n data before the corresponding first history feature value of a data calculates；It is current special by described first
Value indicative shown, first current characteristic value is replaced the characteristic value stored in the default storage unit, to the n into
Row from plus 1 calculate, and return it is described reception data flow in nth data, and from default storage unit obtain before n-1 number
The step of according to corresponding first history feature value.Since total data need not be preserved, according only to current data and history feature value,
The corresponding current characteristic value of total data can be calculated, can not only show current characteristic value in real time, and can be
It largely reduces calculation amount and reduces fortune and deposit.
Based on above-mentioned hardware configuration, the embodiment of data flow real-time processing method of the present invention is proposed.
With reference to Fig. 2, Fig. 2 is the flow diagram of data flow real-time processing method first embodiment of the present invention.
In the first embodiment, the data flow real-time processing method includes the following steps：
Step S10：The nth data in data flow is received, and n-1 data correspond to before acquisition from default storage unit
The first history feature value, the n is integer not less than 1；
It should be noted that the application scenarios of the present embodiment are the data in reception data flow successively, and real-time display is worked as
The corresponding characteristic value of preceding total data.The data flow refers to that primary data can only be read with the sequence provided in advance
One sequence, the default storage unit is for storing history feature value, when needing using the history feature value, by described in
History feature value is read out from the default storage unit.The n is the integer not less than 1, i.e., 1,2,3 ... wait natures
Number.
It is understood that each characteristic value corresponds to a kind of feature, the feature includes the average value of data sample, maximum
Any one of features such as value and minimum value, the First Eigenvalue are the corresponding characteristic value of fisrt feature, and history feature value is preceding n-1
The corresponding characteristic value of a data, the first history feature value are the corresponding history feature value of fisrt feature.For example, described first
It is characterized as average value, when the n is 106, the first history feature value is the corresponding average value of preceding 105 data.Certainly,
When n is 1, the n-1 is 0, and there is no the first history feature values, only perform and receive first data in data flow
Step.
Step S20：N before being calculated according to the nth data and the corresponding first history feature value of the preceding n-1 data
Corresponding first current characteristic value of a data；
It should be noted that current characteristic value is the corresponding characteristic value of preceding n data, first current characteristic value is the
The corresponding current characteristic value of one feature.When the acquisition nth data and corresponding first history feature of the preceding n-1 data
After value, the nth data and the preceding n-1 data corresponding first can be gone through according to the specific nature of fisrt feature
History characteristic value is calculated, corresponding first current characteristic value of n data before obtaining.
For example, when fisrt feature is average value, the n is 21, nth data 10, and the preceding n-1 data correspond to
The first history feature value when being 31, i.e. the 21st number is 10, when the average value of preceding 20 numbers is 31.If using the prior art,
Then need to preserve 21 data before whole, and pass through and be added 21 data before whole, then divided by 21 calculate it is current average
Number.And the method for using the present embodiment, it need to only be calculated using preset formula [n+ (n-1) × μ]/n, specific calculating process
For：The average value of preceding 21 numbers is [10+ (21-1) × 31]/21=30, i.e., described first current characteristic value is 30.The prior art
Total data need to be calculated, when the data volume of total data is huge, fortune is deposited will be full, it is difficult to calculating is continued to run with, and originally
Method need to only calculate two data, and calculation amount is small, and occupancy fortune is deposited low.
Certainly, when n is 1, the n-1 is 0, and there is no the first history feature values, will be directly according to first number
According to the first current characteristic value for calculating previous data.For example, when n is 1, first data is 21, then the first current signature
Be worth is 21.Also, it for different features, carries out calculating first current characteristic value using different preset formulas.
Step S30：First current characteristic value is shown, first current characteristic value is replaced described default
The characteristic value stored in storage unit carries out the n certainly plus 1 calculates, and returns to the nth in the reception data flow
According to, and before being obtained from default storage unit the step of n-1 data corresponding first history feature value.
It is understood that after first current characteristic value is calculated, which is carried out
It has been shown that, so that user obtains first current characteristic value.Meanwhile in order to obtain next first current characteristic value, to the n
It carries out, from adding 1 to calculate, first current characteristic value being replaced the characteristic value stored in the default storage unit, i.e., by preceding n
The corresponding first history feature value of n-1 data before corresponding first current characteristic value of a data is replaced, as calculating next time
The corresponding first history feature value of preceding n data, you can realize and calculate current signature using only current data and history feature value
Value, substantially reduces fortune and deposits.
Further, after the step S30, the method further includes：
The nth data is deleted.
It should be noted that after calculating the first current characteristic value again, only needed due to calculating next current characteristic value
Next data and the first current characteristic value at this time, without current data, will delete current nth data
It removes, so, it is only necessary to preserve two data, just can calculate the characteristic value in data flow, reduce fortune and deposit.
In the present embodiment, by receiving the nth data in data flow, and n-1 before being obtained from default storage unit
The corresponding first history feature value of a data, the n are the integer not less than 1；According to the nth data and the preceding n-1
Corresponding first current characteristic value of n data before the corresponding first history feature value of a data calculates；It is current special by described first
Value indicative shown, first current characteristic value is replaced the characteristic value stored in the default storage unit, to the n into
Row from plus 1 calculate, and return it is described reception data flow in nth data, and from default storage unit obtain before n-1 number
The step of according to corresponding first history feature value.Since total data need not be preserved, according only to current data and history feature value,
The corresponding current characteristic value of total data can be calculated, can not only show current characteristic value in real time, and can be
It largely reduces calculation amount and reduces fortune and deposit.
With reference to Fig. 3, Fig. 3 is the flow diagram of data flow real-time processing method second embodiment of the present invention, based on above-mentioned
Embodiment shown in Fig. 2 proposes the second embodiment of data flow real-time processing method of the present invention.
In a second embodiment, before the step S10, the method further includes：
Step S01：Data processing instructions input by user are received, target signature is extracted from the data processing instructions,
The target signature is the corresponding feature of characteristic value to be shown；
It should be noted that the data processing instructions contain the target signature, the target signature is to be shown
The corresponding feature of characteristic value, including：The features such as average value, minimum value, standard deviation and extreme value.User is being inputted at the data
After reason instruction, the data in data flow will be calculated, and target signature described in real-time display.
Step S03：The classification belonging to the target signature is obtained, when the target signature is the first category feature, is performed
Nth data in the reception data flow, and corresponding first history of n-1 data is special before acquisition from default storage unit
The step of value indicative.
It is understood that before the data in receiving data flow, which kind of spy that will clearly show the data flow needed
Value indicative judges the classification belonging to the target signature.It, need to be to the mesh since the corresponding computing mechanism of each target signature differs
Mark feature classify, according to whether can directly calculate object feature value and can real time implementation classify to target signature,
First category feature includes：Average value, maximum value and minimum value etc. can directly calculate object feature value according to history feature value
Can real time implementation completely feature, it is described real time implementation to refer to input a data that output display in real time one is corresponding immediately completely
Characteristic value.The feature further includes the second category feature, third category feature and the 4th category feature, and second category feature includes mark
Quasi- difference etc. can indirectly be calculated according to history feature value object feature value can real time implementation completely feature, the third category feature packet
The feature for the real time implementation that can be delayed is included, the real time implementation that is delayed refers to after next data are inputted, and side shows that current data corresponds to
Characteristic value, such as extreme value.4th category feature refers to be unable to the data of real time implementation, by this programme to being unable to the data of real time implementation
Real-time display as much as possible.
Further, the step S30, specifically includes：
Using first current characteristic value as first kind characteristic value, and show the first kind characteristic value, described first
Category feature value is the corresponding characteristic value of first category feature, and first current characteristic value is replaced the default storage unit
The characteristic value of middle storage carries out the n from plus 1 calculates, and return to the nth data in the reception data flow, and from pre-
If before being obtained in storage unit the step of n-1 data corresponding first history feature value.
It is to be appreciated that the calculating process of first category feature is relatively simple, it is only necessary to use the first history feature value
Calculate the first current characteristic value, you can using first current characteristic value as first kind characteristic value, exported.
Step S02：Target formula corresponding with the target signature is searched in mapping relations, is wrapped in the mapping relations
Correspondence containing feature and formula.
It should be noted that each target signature has a corresponding calculation formula, and close each feature is corresponding with formula
System is stored in mapping relations.So as to search target formula corresponding with target signature according to target signature in mapping relations,
For calculating first current characteristic value.Also, the correspondence can carry out classification storage according to the classification of target signature.
Further, the step S20, specifically includes：
Step S201：The nth data and the preceding n-1 data corresponding first are gone through according to the target formula
History characteristic value is calculated, and obtains corresponding first current characteristic value of the preceding n data.
It is understood that after target formula is found according to target signature, will be calculated according to the target formula
First current characteristic value.
In the present embodiment, it by receiving data processing instructions input by user, is extracted from the data processing instructions
Target signature；Corresponding with target signature target formula is searched in mapping relations, according to target formula calculating described the
One current characteristic value obtains the classification belonging to the target signature, when the target signature is the first category feature, described in execution
Receive data flow in nth data, and from default storage unit obtain before the corresponding first history feature value of n-1 data
The step of.Due to obtaining the calculating demand of user, the corresponding target formula of characteristic value to be shown may correspondingly determine that, so as to
Quickly calculated.
With reference to Fig. 4, Fig. 4 is the flow diagram of data flow real-time processing method 3rd embodiment of the present invention, based on above-mentioned
Embodiment shown in Fig. 3 proposes the 3rd embodiment of data flow real-time processing method of the present invention.
In the present embodiment, after the step S02, the method further includes：
Step S04：When the target signature is the second category feature, the nth data in the data flow is received, and from
The corresponding second history feature value of n-1 data before being obtained in the default storage unit；
It is understood that second category feature can calculate mesh indirectly including standard deviation etc. according to the second history feature value
Mark characteristic value can real time implementation completely feature.The Second Eigenvalue be the corresponding characteristic value of second feature, history feature value
For the corresponding characteristic value of preceding n-1 data, the second history feature value is the corresponding history feature value of second feature.For example,
Second category feature is standard deviation, determines that the corresponding second feature of standard deviation includes root mean square according to the preset algorithm of standard deviation
And average value, when the n is 106, the second history feature value is the corresponding root mean square of preceding 105 data and average value.
Certainly, when n is 1, the n-1 is 0, and there is no the second history feature values, only perform first received in data flow
The step of data.
Step S05：N before being calculated according to the nth data and the corresponding second history feature value of the preceding n-1 data
The corresponding second category feature value of a data, the second category feature value are the corresponding characteristic value of second category feature；
It should be noted that current characteristic value is the corresponding characteristic value of preceding n data, second current characteristic value is the
The corresponding current characteristic value of two features.When the acquisition nth data and corresponding second history feature of the preceding n-1 data
After value, the nth data and the preceding n-1 data corresponding second can be gone through according to the specific nature of second feature
History characteristic value is calculated, the corresponding second category feature value of n data before obtaining.
Step S06：The second category feature value is shown, the second history feature value is updated, to institute
N is stated to carry out adding 1 to calculate certainly, and return to the nth data received in the data flow, and obtain from default storage unit
The step of preceding n-1 data corresponding second history feature value.
It is understood that after the second category feature value is calculated, which is shown,
So that user obtains the second category feature value.Meanwhile in order to obtain next second category feature value, the n is carried out to add 1 certainly
It calculates, second current characteristic value is replaced to the characteristic value stored in the default storage unit, i.e., correspond to preceding n data
The second current characteristic value replace before the corresponding second history feature value of n-1 data, as next time calculating preceding n data pair
The the second history feature value answered, you can realize and calculate current goal characteristic value using only current data and history feature value, significantly
Fortune is reduced to deposit.
Further, the step S05, specifically includes：
Step S051：The nth data and the corresponding second history feature value of the preceding n-1 data are counted
It calculates, corresponding second current characteristic value of n data before obtaining；
Step S052：The preceding n data pair are calculated according to corresponding second current characteristic value of the preceding n data
The the second category feature value answered.
It should be noted that the derivation formula due to standard deviation isIt is only capable of indirectly by Second Eigenvalue
It obtains the second category feature value, i.e., the second current characteristic value need to be calculated by the second history feature value, then current by second
The second category feature value is calculated in characteristic value.For example, the calculating process of standard deviation includes：Pass through the history of root mean square, average value
The current characteristic value of root mean square, average value is calculated in characteristic value, passes through the current spy of root mean square, average value according to preset formula
Current standard deviation is calculated in value indicative.
Further, described in the step S06 is updated the second history feature value, specifically includes：
Second current characteristic value is replaced to the second history feature value stored in the default storage unit.
It is illustrated below using second category feature as standard deviation.
Standard deviation formula：
Root mean square formula：
Mean Value Formulas：
Root mean square formula and Mean Value Formulas is brought into standard deviation formula to obtain：
Wherein, σ is standard deviation,For root mean square, μ is average value, and n is sample data number, and Xi is i-th of data.
Therefore, the 1st data X is inputted1Afterwards, it can calculateAnd μ1；Input nth data Xn-1Afterwards, it calculatesWith
μn-1, and calculate σn-1.Then as input nth data XnAfterwards, μ is calculatedn=[(n-1) μn-1+Xn]/n、
So as to calculate in real time
σ is calculated every time2And open radical sign, you can real-time outputting standard difference σ.All input datas need not then be cached
Come and without all data are done the operation such as cumulative again.
For example, it is 13 that the n, which is the 10, the 10th data, the average value 8 of preceding 9 numbers, during root mean square 8.18.If using existing
There is technology, then 10 data before preservation whole are needed, and by calculating average value to 10 data before whole, further according to standard deviation
Formula calculate standard deviation, calculation times are more, occupy fortune deposits it is big.And the method for using the present embodiment, only it need to use preset formula
It is calculated, specific calculating process is：The average value of preceding 10 numbers be (13+8 × 9) ÷ 10=8.5, the root mean square of preceding 10 numbers
ForFormula according to being derived from is calculated, and the corresponding standard deviation of preceding 10 numbers isThe prior art need to calculate total data, when the data volume of total data is huge,
Fortune is deposited will be full, it is difficult to calculating is continued to run with, and this method need to only calculate three data, calculation amount is small, and occupancy fortune is deposited low.
In the present embodiment, when the target signature is the second category feature, the nth in the data flow is received
According to, and from the default storage unit obtain before the corresponding second history feature value of n-1 data；According to the nth
The corresponding second category feature value of n data before being calculated according to the second history feature value corresponding with the preceding n-1 data, described the
Two category feature values are the corresponding characteristic value of second category feature；The second category feature value is shown, to described second
History feature value is updated, and the n is carried out certainly plus 1 calculates, and return to the nth received in the data flow
According to, and before being obtained from default storage unit the step of n-1 data corresponding second history feature value.Due to complete without preserving
According only to current data and history feature value, the corresponding current characteristic value of total data can be calculated in portion's data, can not only
It is enough to show current characteristic value in real time, and largely reduce calculation amount and reduce fortune and deposit.
With reference to Fig. 5, Fig. 5 is the flow diagram of data flow real-time processing method fourth embodiment of the present invention, based on above-mentioned
Embodiment shown in Fig. 4 proposes the fourth embodiment of data flow real-time processing method of the present invention.
In the present embodiment, after the step S02, the method further includes：
Step S07：When the target signature is third category feature, the nth data in the data flow and n-th are received
+ 1 data, and the third history for obtaining from the default storage unit (n-1)th data and the nth data is special
Value indicative；
It is understood that for the third category feature value for the real time implementation that can be delayed, due to the feature of the third category feature,
It is difficult to export object feature value completely in real-time, such as extreme value, after next data need to be inputted, data can be calculated, obtained
Go out whether current data is extreme value.Third feature value is the corresponding characteristic value of third feature, when third category feature value is extreme value,
Third history feature value is (n-1)th data, the corresponding comparison result of nth data.For example, when the n is 106, it is described
Third history feature value is the 105th data and the 106th corresponding comparison result of data.Certainly, when n is 1, the n-1
It is 0, there is no the third history feature values, only perform the step of receiving first data in data flow.
Step S08：The third current characteristic value of the nth data and (n+1)th data is calculated, by the third
History feature value is compared with the third current characteristic value, and third category feature value, the third are obtained according to comparison result
Category feature value is the corresponding characteristic value of the third category feature；
It should be noted that the third current characteristic value is nth data and the corresponding relatively knot of (n+1)th data
Fruit.It, can be according to third class after the nth data and the preceding n-1 data corresponding third history feature value is obtained
The specific nature of feature calculates the nth data and the corresponding third history feature value of the preceding n-1 data,
The corresponding third current characteristic value of n data before obtaining.Certainly, when n is 1, the n-1 is 0, and there is no the third history
Characteristic value will directly calculate the third current characteristic value of previous data according to first data.For example, when n is 1, first
A data are 21, and second data are 20, then third current characteristic value be more than.
Step S09：The third category feature value is shown, the third current characteristic value is replaced into described preset and is deposited
The third history feature value stored in storage unit carries out the n certainly plus 1 calculates, and returns to n-th received in the data flow
The step of a data and (n+1)th data.
It is understood that after the third category feature value is calculated, which is shown,
So that user obtains the third category feature value.Meanwhile in order to obtain next third category feature value, the n is carried out to add 1 certainly
It calculates, the third current characteristic value is replaced to the characteristic value stored in the default storage unit, i.e., correspond to preceding n data
Third current characteristic value replace before the corresponding third history feature value of n-1 data, as next time calculating preceding n data pair
The third history feature value answered, you can realize and calculate current characteristic value using only current data and history feature value, greatly reduce
Fortune is deposited.
It is illustrated below using the third category feature as extreme value.
In the concrete realization, X is preservedn-1And XnBetween comparison symbol, when receiving Xn+1When, obtain XnAnd Xn+1Between
Compare symbol, two comparison symbols are compared, X is judged according to comparison resultnWhether it is extreme value.When two comparison symbol phases
Meanwhile illustrate XnIt is not extreme value；When two comparison symbol differences, the position N of extreme point and extreme value X is exportedn。
Further, after the step S02, the method further includes：
When the target signature is four category feature, the n-th second data, the n-th+0.5 number of seconds in the data flow are received
According to and the (n+1)th second data；
Default processing is carried out to n-th second data, the n-th+0.5 second data and the (n+1)th second data；
The intermediate data of the n-th+0.5 second data after intercepting process, will be in interception according to the interception part corresponding time
Between data replace to the data among treated the n-th second data and the (n+1)th second data；
The replaced current data of real-time display carries out the n certainly plus 1 calculates, and returns and receive in the data flow
The n-th second data, the n-th+0.5 second data and the step of the (n+1)th second data；
Wherein, n-th second data is the data of 1 second since n-th second, and such as the 1.5th second data is opened from the 1.5th second
The data of 1 second to begin.
It should be noted that when the 4th category feature is noise reduction data, the n-th second data of the reception, the n-th+0.5 second
Data and the (n+1)th second data are that band is made an uproar data, and the default processing refers to noise reduction process, and the current data is noise reduction number
According to.
It is illustrated so that the 4th category feature is the data block formula wave filter of global type as an example below in conjunction with Fig. 6.Such as
Described in Fig. 6, Fig. 6 is noise reduction process flow diagram.When n is 1, the 1st second data, the 1.5th second data in data flow are received
And the 2nd second data, intercept the middle section of the 1.5th second data, i.e. the 1.75th second data, according to the 1.75th second data start-stop when
Between the tail portion of the 1st second data of corresponding interception and the head of the 2nd second data, the 1.75th second data is spliced to the tail of the 1st second data
In portion and the head of the 2nd second data, you can obtain the smooth noise reduction data of the 1st second data to the 2nd second data.By noise reduction data
Output in real time, you can constantly output band is made an uproar the corresponding noise reduction data of data in real time.
In the present embodiment, by when the target signature is third category feature, receiving n-th in the data flow
Data and (n+1)th data, compare and obtain third category feature value, can postpone to receive the time real-time display third class of a data
Characteristic value.By when the target signature is four category feature, receive the n-th second data in the data flow, the n-th+0.5 second
Data and the (n+1)th second data；Processing data are exported by replacing.Since total data need not be preserved, according only to current data and
The corresponding object feature value of total data can be calculated in history feature value, can not only display target characteristic value in real time,
It largely reduces calculation amount and reduces fortune and deposit, and can realize the processing data of real-time output smoothing.
In addition, the embodiment of the present invention also proposes a kind of storage medium, be stored with data flow on the storage medium locates in real time
Program is managed, following operation is realized when the data flow real-time handler is executed by processor：
The nth data in data flow is received, and n-1 data corresponding first are gone through before acquisition from default storage unit
History characteristic value, the n are the integer not less than 1；
N data pair before being calculated according to the nth data and the corresponding first history feature value of the preceding n-1 data
The first current characteristic value answered；
First current characteristic value is shown, first current characteristic value is replaced into the default storage unit
The characteristic value of middle storage carries out the n from plus 1 calculates, and return to the nth data in the reception data flow, and from pre-
If before being obtained in storage unit the step of n-1 data corresponding first history feature value.
Further, following operation is also realized when the data flow real-time handler is executed by processor：
Data processing instructions input by user are received, target signature, the target are extracted from the data processing instructions
It is characterized as the corresponding feature of characteristic value to be shown；
The classification belonging to the target signature is obtained, when the target signature is the first category feature, performs the reception
Nth data in data flow, and from default storage unit obtain before the corresponding first history feature value of n-1 data step
Suddenly；
Correspondingly, it is described to show first current characteristic value, it specifically includes：
Using first current characteristic value as first kind characteristic value, and show the first kind characteristic value, described first
Category feature value is the corresponding characteristic value of first category feature.
Further, following operation is also realized when the data flow real-time handler is executed by processor：
Search corresponding with target signature target formula in mapping relations, in the mapping relations comprising feature and
The correspondence of formula；
Correspondingly, it is described to be calculated according to the nth data with the corresponding first history feature value of the preceding n-1 data
Corresponding first current characteristic value of preceding n data, specifically includes：
According to the target formula to the nth data and the corresponding first history feature value of the preceding n-1 data
It is calculated, obtains corresponding first current characteristic value of the preceding n data.
Further, following operation is also realized when the data flow real-time handler is executed by processor：
When the target signature is the second category feature, the nth data in the data flow is received, and preset from described
The corresponding second history feature value of n-1 data before being obtained in storage unit；
N data pair before being calculated according to the nth data and the corresponding second history feature value of the preceding n-1 data
The the second category feature value answered, the second category feature value are the corresponding characteristic value of second category feature；
The second category feature value is shown, the second history feature value is updated, the n is carried out certainly
Add 1 calculating, and return to the nth data received in the data flow, and n-1 number before being obtained from default storage unit
The step of according to corresponding second history feature value.
Further, following operation is also realized when the data flow real-time handler is executed by processor：
The nth data and the corresponding second history feature value of the preceding n-1 data are calculated, obtain preceding n
Corresponding second current characteristic value of a data；
The preceding n data corresponding second are calculated according to corresponding second current characteristic value of the preceding n data
Category feature value.
Further, following operation is also realized when the data flow real-time handler is executed by processor：
Second current characteristic value is replaced to the second history feature value stored in the default storage unit.
Further, following operation is also realized when the data flow real-time handler is executed by processor：
The nth data is deleted.
Further, following operation is also realized when the data flow real-time handler is executed by processor：
When the target signature is third category feature, the nth data and (n+1)th number in the data flow are received
According to, and obtain from the default storage unit third history feature value of (n-1)th data and the nth data；
The third current characteristic value of the nth data and (n+1)th data is calculated, by the third history feature
Value is compared with the third current characteristic value, and third category feature value, the third category feature value are obtained according to comparison result
For the corresponding characteristic value of the third category feature；
The third category feature value is shown, the third current characteristic value is replaced in the default storage unit
The third history feature value of storage carries out the n from plus 1 calculates, and return receive nth data in the data flow and
The step of (n+1)th data.
In the present embodiment, by receiving the nth data in data flow, and n-1 before being obtained from default storage unit
The corresponding first history feature value of a data, the n are the integer not less than 1；According to the nth data and the preceding n-1
Corresponding first current characteristic value of n data before the corresponding first history feature value of a data calculates；It is current special by described first
Value indicative shown, first current characteristic value is replaced the characteristic value stored in the default storage unit, to the n into
Row from plus 1 calculate, and return it is described reception data flow in nth data, and from default storage unit obtain before n-1 number
The step of according to corresponding first history feature value.Since total data need not be preserved, according only to current data and history feature value,
The corresponding current characteristic value of total data can be calculated, can not only show current characteristic value in real time, and can be
It largely reduces calculation amount and reduces fortune and deposit.
It should be noted that herein, term " comprising ", "comprising" or its any other variant are intended to non-row
His property includes, so that process, method, article or system including a series of elements not only include those elements, and
And it further includes other elements that are not explicitly listed or further includes intrinsic for this process, method, article or system institute
Element.In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded that including this
Also there are other identical elements in the process of element, method, article or system.
The embodiments of the present invention are for illustration only, do not represent the quality of embodiment.
The use of word first, second, and third does not indicate that any sequence, can these words be construed to title.
Through the above description of the embodiments, those skilled in the art can be understood that above-described embodiment side
Method can add the mode of required general hardware platform to realize by software, naturally it is also possible to by hardware, but in many cases
The former is more preferably embodiment.Based on such understanding, technical scheme of the present invention substantially in other words does the prior art
Going out the part of contribution can be embodied in the form of software product, which is stored in a storage medium
In (such as ROM/RAM, magnetic disc, CD), used including some instructions so that a station terminal equipment (can be mobile phone, computer takes
Be engaged in device, air conditioner or the network equipment etc.) perform method described in each embodiment of the present invention.
It these are only the preferred embodiment of the present invention, be not intended to limit the scope of the invention, it is every to utilize this hair
The equivalent structure or equivalent flow shift that bright specification and accompanying drawing content are made directly or indirectly is used in other relevant skills
Art field, is included within the scope of the present invention.
Claims (10)
1. a kind of data flow real-time processing method, which is characterized in that the described method comprises the following steps：
The nth data in data flow is received, and corresponding first history of n-1 data is special before acquisition from default storage unit
Value indicative, the n are the integer not less than 1；
N data are corresponding before being calculated according to the nth data and the corresponding first history feature value of the preceding n-1 data
First current characteristic value；
First current characteristic value is shown, first current characteristic value is replaced in the default storage unit and is deposited
The characteristic value of storage carries out the n certainly plus 1 calculates, and returns to the nth data in the reception data flow, and deposit from default
Before being obtained in storage unit the step of n-1 data corresponding first history feature value.
2. data flow real-time processing method as described in claim 1, which is characterized in that n-th received in data flow
Data, and before being obtained from default storage unit before the corresponding first history feature value of n-1 data, the method is also wrapped
It includes：
Data processing instructions input by user are received, target signature, the target signature are extracted from the data processing instructions
For the corresponding feature of characteristic value to be shown；
The classification belonging to the target signature is obtained, when the target signature is the first category feature, performs the reception data
Nth data in stream, and before being obtained from default storage unit the step of n-1 data corresponding first history feature value；
Correspondingly, it is described to show first current characteristic value, it specifically includes：
Using first current characteristic value as first kind characteristic value, and show the first kind characteristic value, the first kind is special
Value indicative is the corresponding characteristic value of first category feature.
3. data flow real-time processing method as claimed in claim 2, which is characterized in that described to obtain belonging to the target signature
Classification, in the target signature for after the first category feature, the method further includes：
Target formula corresponding with the target signature is searched in mapping relations, feature and formula are included in the mapping relations
Correspondence；
Correspondingly, it is described according to the nth data and the corresponding first history feature value of the preceding n-1 data calculate before n
Corresponding first current characteristic value of a data, specifically includes：
The nth data and the corresponding first history feature value of the preceding n-1 data are carried out according to the target formula
It calculates, obtains corresponding first current characteristic value of the preceding n data.
4. data flow real-time processing method as claimed in claim 2, which is characterized in that described to obtain belonging to the target signature
Classification after, the method further includes：
When the target signature is the second category feature, the nth data in the data flow is received, and from the default storage
The corresponding second history feature value of n-1 data before being obtained in unit；
N data are corresponding before being calculated according to the nth data and the corresponding second history feature value of the preceding n-1 data
Second category feature value, the second category feature value are the corresponding characteristic value of second category feature；
The second category feature value is shown, the second history feature value is updated, the n is carried out to add 1 certainly
It calculates, and returns to the nth data received in the data flow, and n-1 data before being obtained from default storage unit
The step of corresponding second history feature value.
5. data flow real-time processing method as claimed in claim 4, which is characterized in that it is described according to the nth data and
The corresponding second category feature value of n data before the corresponding second history feature value of the preceding n-1 data calculates, specifically includes：
The nth data and the corresponding second history feature value of the preceding n-1 data are calculated, n number before obtaining
According to corresponding second current characteristic value；
It is special that corresponding second class of the preceding n data is calculated according to corresponding second current characteristic value of the preceding n data
Value indicative.
6. data flow real-time processing method as claimed in claim 5, which is characterized in that described to the second history feature value
It is updated, specifically includes：
Second current characteristic value is replaced to the second history feature value stored in the default storage unit.
7. data flow real-time processing method as described in claim 1, which is characterized in that described to return in the reception data flow
Nth data, and before being obtained from default storage unit the step of n-1 data corresponding first history feature value after,
The method further includes：
The nth data is deleted.
8. data flow real-time processing method as claimed in claim 2, which is characterized in that described to obtain belonging to the target signature
Classification after, the method further includes：
When the target signature is third category feature, the nth data and (n+1)th data in the data flow are received, and
The third history feature value of (n-1)th data and the nth data is obtained from the default storage unit；
Calculate the third current characteristic value of the nth data and (n+1)th data, by the third history feature value with
The third current characteristic value is compared, and obtains third category feature value according to comparison result, the third category feature value is institute
State the corresponding characteristic value of third category feature；
The third category feature value is shown, the third current characteristic value is replaced in the default storage unit and is stored
Third history feature value, the n is carried out from plus 1 calculated, and return receive nth data in the data flow and n-th+
The step of 1 data.
9. a kind of data flow real-time processing device, which is characterized in that described device includes：Memory, processor and it is stored in institute
State the data flow real-time handler that can be run on memory and on the processor, the data flow real-time handler quilt
The step of data flow real-time processing method as described in claim 1 to 8 is realized when the processor performs.
10. a kind of storage medium, which is characterized in that data flow real-time handler, the number are stored on the storage medium
The step of the data flow real-time processing method as described in claim 1 to 8 is realized when being executed by processor according to stream real-time handler
Suddenly.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711274270.1A CN108153591A (en) | 2017-12-05 | 2017-12-05 | Data flow real-time processing method, device and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711274270.1A CN108153591A (en) | 2017-12-05 | 2017-12-05 | Data flow real-time processing method, device and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108153591A true CN108153591A (en) | 2018-06-12 |
Family
ID=62465978
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711274270.1A Pending CN108153591A (en) | 2017-12-05 | 2017-12-05 | Data flow real-time processing method, device and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108153591A (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101059802A (en) * | 2006-04-17 | 2007-10-24 | 中国科学院计算技术研究所 | Method and system for calculating data flow maximum value and minimum value under sliding window |
| US8650167B2 (en) * | 2004-09-17 | 2014-02-11 | Dell Software Inc. | Method and system for data reduction |
| CN104063747A (en) * | 2014-06-26 | 2014-09-24 | 上海交通大学 | Performance abnormality prediction method in distributed system and system |
| CN104090952A (en) * | 2014-07-02 | 2014-10-08 | 华中科技大学 | Method and system for estimating average value of data flow under sliding window |
| CN104484673A (en) * | 2014-12-05 | 2015-04-01 | 南京大学 | Data complementation method for pattern recognition application of real-time data flow |
| CN105868026A (en) * | 2016-04-05 | 2016-08-17 | 浪潮电子信息产业股份有限公司 | Method and device for calculating sequence average value |
-
2017
- 2017-12-05 CN CN201711274270.1A patent/CN108153591A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8650167B2 (en) * | 2004-09-17 | 2014-02-11 | Dell Software Inc. | Method and system for data reduction |
| CN101059802A (en) * | 2006-04-17 | 2007-10-24 | 中国科学院计算技术研究所 | Method and system for calculating data flow maximum value and minimum value under sliding window |
| CN104063747A (en) * | 2014-06-26 | 2014-09-24 | 上海交通大学 | Performance abnormality prediction method in distributed system and system |
| CN104090952A (en) * | 2014-07-02 | 2014-10-08 | 华中科技大学 | Method and system for estimating average value of data flow under sliding window |
| CN104484673A (en) * | 2014-12-05 | 2015-04-01 | 南京大学 | Data complementation method for pattern recognition application of real-time data flow |
| CN105868026A (en) * | 2016-04-05 | 2016-08-17 | 浪潮电子信息产业股份有限公司 | Method and device for calculating sequence average value |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10783151B1 (en) | Popularity-based content feed management system | |
| US20210026909A1 (en) | System and method for identifying contacts of a target user in a social network | |
| US20170161645A1 (en) | Method and apparatus for labeling training samples | |
| US10970554B2 (en) | Method and system for automatically producing video highlights | |
| CN111898683A (en) | Image classification method and device based on deep learning and computer equipment | |
| WO2019184640A1 (en) | Indicator determination method and related device thereto | |
| CN108234347A (en) | A kind of method, apparatus, the network equipment and storage medium for extracting feature string | |
| CN109101658A (en) | Information search method, device and equipment/terminal/server | |
| CN106528894B (en) | The method and device of label information is set | |
| CN105574030B (en) | A kind of information search method and device | |
| EP2389624A1 (en) | Sampling analysis of search queries | |
| CN113011581A (en) | Neural network model compression method and device, electronic equipment and readable storage medium | |
| CN109618236A (en) | Video comments treating method and apparatus | |
| CN114726823B (en) | Domain name generation method, device and equipment based on generation countermeasure network | |
| CN107465886A (en) | Intermediary method, mediating device and system and record media | |
| CN109922379A (en) | Advertisement video optimization method, device, equipment and computer readable storage medium | |
| CN108764455A (en) | Parameter adjustment method, device and storage medium | |
| US20220124185A1 (en) | Terabit-scale network packet processing via flow-level parallelization | |
| CN109739989A (en) | File classification method and computer equipment | |
| CN103324679B (en) | Control method and equipment that in a kind of caching server, data update | |
| CN110969240A (en) | Pruning method, device, equipment and medium for deep convolutional neural network | |
| CN111737473A (en) | Text classification method, device and equipment | |
| JP2011209800A (en) | Privacy protection device for disclosed information, privacy protection method for the disclosed information and program | |
| CN108153591A (en) | Data flow real-time processing method, device and storage medium | |
| CN105589896B (en) | Data digging method and device |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180612 |