CN110428333A - For the Data Detection and analysis method of financial intermediary MT4 foreign exchange transaction software - Google Patents
For the Data Detection and analysis method of financial intermediary MT4 foreign exchange transaction software Download PDFInfo
- Publication number
- CN110428333A CN110428333A CN201910706060.8A CN201910706060A CN110428333A CN 110428333 A CN110428333 A CN 110428333A CN 201910706060 A CN201910706060 A CN 201910706060A CN 110428333 A CN110428333 A CN 110428333A
- Authority
- CN
- China
- Prior art keywords
- market
- data
- index
- time
- transaction
- 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
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2455—Query execution
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/04—Trading; Exchange, e.g. stocks, commodities, derivatives or currency exchange
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Computational Linguistics (AREA)
- Databases & Information Systems (AREA)
- General Engineering & Computer Science (AREA)
- Development Economics (AREA)
- Economics (AREA)
- Marketing (AREA)
- Strategic Management (AREA)
- Technology Law (AREA)
- General Business, Economics & Management (AREA)
- Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)
Abstract
The invention discloses a kind of Data Detections and analysis method for financial intermediary MT4 foreign exchange transaction software, and content includes S1: firm offer opens an account and injects a large amount of true funds；S2: creation unification and the high computer techno-stress operation hardware platform matched；S3: science writes trading algorithms；S4: data import Cloud Server；S5: for the monitoring of market, the market abnormality judgment criterions obtained according to the neural network model that step S3 is created are divided into situations such as market are interrupted, quotations is abnormal, by diagnostic result real-time Transmission to alarm center, notice is manually handled or is intervened, and more serious consequence is avoided.S6: this four exponent datas of binding site difference environmental index, market environmental index, cost index overnight, trading environment index, carry out a comprehensive scoring analysis, finally obtain a composite index, by checking this composite index, user can intuitively tell the superiority and inferiority of each broker's platform.
Description
Technical field
The present invention relates to the Internet transmission technology and finance data analysis field, it is especially a kind of based on big data excavate
Diagnose the monitoring method of the MT4 platform superiority and inferiority of each foreign-exchange broker.
Background technique
With the continuous heating of public investment enthusiasm, recent domestic remittance industry has welcome fast development, in addition to some
Outside time-honored old brand foreign-exchange broker, numerous new foreign-exchange brokers emerge rapidly in large numbersBamboo shoots after a spring rain as occur, wherein be no lack of it is some after rise
Show, but also have and much adulterate, or even very much " black platforms " occur.Numerous industry participants enter market, and channel stands in great numbers,
Dispersion phenomenon is still more serious.Important participation tool of the transaction platform as investor, differentiation are also quite serious.
A workman wants to do his work well must first sharpen his tools.For investor, regardless of any investment, what is most thirsted for is to possess a public affairs
Flat, investment environment openly, just.Some " evil mind " broker or are set in software by manipulating to platform software backstage
Many transaction obstacles are set, sink into the transaction of investor among unfavorable.Even if you have market judgement
Mostly precisely, still it is unable to escape the destiny gathered in by broker.Occurred the case that many black platforms are exposed in recent years, it is vertical to have very much
It invests " veteran ", also still can not clearly distinguish the true and false and existing various problems of these platforms, still more many just introductions
New hand, even more feel at a loss.
For common investor, the superiority and inferiority of a platform is distinguished, be more self-dependent experience, or by certainly
Some transaction of body carry out rough analysis.But on the one hand the number of transaction of individual is limited, small data is not big reference
Meaning does not have representativeness；Another aspect individual completely can not be monitored and be analyzed to all periods, all conclusions
It is all unilateral some analyses.It therefore is totally impossible something for individual for the superiority and inferiority of discrimination platform.By
Naturally there is opacity in the market OTC, wherein the 60% of client trading loss is from hiding transaction cost.Market is pair
Rush the institutional clients such as fund, principal mark and high frequency transaction and provide solution, however, even to this day, general transaction person and these
Solution is still out of reach.
For each platform of existing market status that the good and bad jumbled together, investor also recognizes this serious problems, but incompetent
For power, can use a good platform can only lean on fortune.On the market, some software developers have also been made one for this problem
, often there is larger prolong on quotations research in the small tool detected a bit, the local sampling analysis that traditional software quotient uses
Late, and data packetloss is serious, and the analysis conclusion obtained is difficult to reflect the truth of platform strictly according to the facts.Otherwise for domestic thousands of
Family broker, strength and energy aspect are unable to monitor out trading environment data, this just has a greatly reduced quality to the effect of evaluation and test.These works
Tool is often all embarrassed to be used greatly, is helped without any essence investor, slowly but also has been developed some broker being had ulterior motives
The tool utilized is more degrading this market, and to this, chaotic trade market environment has undoubtedly made the matter worse.
Summary of the invention
In view of above-mentioned technical problem, the present invention provides a kind of comprehensively and effectively MT4 foreign exchange transaction software integrationizations to monitor
Scheme, by using a series of internet information data processings for each financial intermediary MT4 foreign exchange transaction software with
Analysis tells the superiority and inferiority of each financial broker's platform for investor, then provides a transparent transaction city for investor
.
In order to achieve the above objectives, technical solution of the present invention is:
For the Data Detection and analysis method of financial intermediary MT4 foreign exchange transaction software, specifically includes the following steps:
S1: the financial broker more than any one family to be monitored progress firm offer is opened an account and is infused at home using internet
Enter a large amount of true funds, all Account Types opened an account all are unified standard account；
S2: to ensure to monitor the just of environment, unification is created for the MT4 software of all financial broker to be monitored
And high computer techno-stress operation hardware platform (being configured to CPU: eight core 2.5GHz, memory: 32GB, bandwidth: 100M) matched；
S3: writing trading algorithms, the input data format of Recognition with Recurrent Neural Network model is defined, for the difference of foreign exchange market
Disk constructs Trading Model, and resolves the period for non-agricultural period and interest rate, carries out real-time monitoring to market, when monitoring market
Uniformly placed an order transaction when big ups and downs；
S4: data import Cloud Server, i.e. mass data is put in storage, in terms of market, to each of each monitoring object
The data of pen have carried out comprehensive record, local zone time, server time, buying price, selling price, newest valence, point comprising this data
The whole quotation information such as difference, in addition generate 1 minute, 5 minutes, 30 minutes, 1 hour, 4 hours, 1 day, 1 week in real time automatically again etc.
Multicycle data, this cycle data are used for the detection processing in later period；
It in terms of transaction, is traded by sequencing, realizes that machine completion automated intelligent places an order, the data in MT4 are passed
The long short-term memory layer for transporting to Recognition with Recurrent Neural Network model, carries out the training of neural network model, passes through the training method of system
With the parameter in adaptive estimation transaction, screens judge special market (including common market, high frequency market etc.) in real time, model of trading
It encloses and covers all constant time ranges, guarantee the integrality and high efficiency of evaluation and test data as far as possible；By the local zone time of each order, clothes
The business device time type, lower single quantity, lower single direction, lower single price, allows sliding points, stops loss valence, being only full of valence, server and close a position
Time, local close a position time, settlement price, quantity of closing a position, the whole storages in real time of all information such as sliding point of closing a position；And it is all
Sheet information is all put in storage comprising all sheet types such as BUYLIMIT, SELLLIMIT, BUYSTOP, SELLSTOP, is established huge
Big Database Systems design rigorous data dictionary, formulate specific data number, real-time tracking data path, each family is all
Strict classification is truly realized data mostly without unrest；
S5: for the monitoring of market, the market abnormal determination mark obtained according to the neural network model that step S3 is created
Standard is divided into market interruption, the abnormal two kinds of situations of quotations, by diagnostic result real-time Transmission to alarm center, notice is artificial
It is handled or is intervened, avoid more serious consequence；
S6: this four indexes of binding site difference environmental index, market environmental index, cost index overnight, trading environment index
Data carry out a comprehensive scoring analysis, and different weights is arranged in the influence situation according to each single item in practice to transaction, most
A composite index is obtained eventually, and historical data finally is gathered according to the difference in period to these data again, generates refer to day respectively
Number, all indexes and monthly index, by checking this composite index, user can intuitively tell the excellent of each broker's platform
It is bad.
In above scheme, trading algorithms use the dynamic analysis to market in the step S3, in the different tendencies of market
Lower such as concussion market, unilateral market all carry out the transaction that places an order, and accomplish the accurate transaction of each market of day part.
In above scheme, the Trading Model in the step S3 can obtain in real time market.
In above scheme, the market for the lower broker's platform of some market frequencies are obtained in real time using efficient
EA scripting language, can accomplish Millisecond delay acquisition.
It is described to use multithreading dual-mode network model for frequent broker's platform of offering in above scheme, make market
It monitors error time and is no more than 10 milliseconds, to realize the seamless monitoring of annual all the period of time to market.
In above scheme, the alarm center can receive monitoring data information in real time, be encoded according to the market received
And timestamp, intelligent Matching abnormal conditions, and addressee is notified in a manner of mail or short message etc. in time, to reach timely processing
Purpose.
In above scheme, the alarm center is provided with multiple Data Nodes, can carry out flexible configuration, Ke Yitong to it
When be sent to multiple addressees, and different contents of announcement can also be set for different addressees, have safely controllable
And no-delay feature.
It is described to be directed in method to the real time monitoring in terms of market in above scheme, it is provided with marking scheme, the program pair
Quotation frequency, the server reconnection time, puts poor maximum value, point difference minimum value, puts poor expansion, more simple interests, empty list broken string number
Interest, more empty interests and these data obtain every exponential number respectively, then complete full depth by calculating variance yields
It calculates, finally obtains market environmental index, puts poor environmental index, cost index overnight；Wherein the index value full marks of each single item are
10 points, the superiority and inferiority of the platform market environment can be intuitively told by these calculated numerical value.
In above scheme, to transaction most jogging speed, transaction prestissimo, transaction average speed in the method；Order is just
The negative sliding point of sliding point, order and sliding point these data of probability of generation derived, calculate by pen according to covariance, generate final
Trading environment index.
The method of the present invention has the beneficial effect that:
1, make investment that can clearly distinguish each broker's platform superiority and inferiority, a protection work well is played to the investment of itself
With.In terms of market monitoring, the data such as exchange quotations data difference domestic equity futures, because foreign exchange market global investor is all
There are participation, substantial amounts, therefore price change is quite active.Market monitor system and carry out panorama to all foreign-exchange broker platforms
Formula scanning, can real-time capture arrive each data, by cloud handle in real time analyze and generate accordingly report result.User is not
Each platform software must be opened and carry out observation comparison, by checking that these data can be very clear to all platforms.It hands over
In terms of easy environmental monitoring, realization covers all broker's platforms comprehensively, parses All Activity order, suspicious order is allowed nowhere to escape
Shape can be used family and easily see clearly that who on earth is being swum nakedly.
2, critical facilitation is played to the benign development of foreign exchange market.Foreign exchange market has its own reason miscellaneous always
Disorderly without chapter, lack an effective supervision all the time.Data through the invention evaluate and test analysis report, eliminate the false and retain the true, can be with
So that this market is become further transparent, each continuous rigors of platform broker oneself can be more effectively supervised in turn, by not
It is disconnected to optimize itself transaction platform performance, better service is provided, more wide market only in this way can be just possessed.
Specific embodiment
In order to be easy to understand the technical means, the creative features, the aims and the efficiencies achieved by the present invention, tie below
Specific embodiment is closed, the present invention is further explained.
The present embodiment is directed to the Data Detection and analysis method of financial intermediary MT4 foreign exchange transaction software, as one kind
Depth network neural algorithm based on cloud big data, specifically includes the following steps:
Step 1: thousands of broker's progress firm offers open an account and inject a large amount of true funds at home, and Account Type is all system
One standard account.Purpose is to monitor MT4 trading environment and market environment under unified Account Type.
Step 2: selecting top server environment to configure, using uniform hardware environment, including CPU, memory, network bandwidth etc.
(general use is configured to CPU: eight cores, 2.5GHz, memory: 32GB, bandwidth: 100M here), installs what each broker provided
MT4 software makes every effort to guarantee that the height of monitoring environment is outstanding, and makes the MT4 software of each broker all in same software and hardware
In environment, guarantee the fairness of monitoring environment.
Step 3: science writes trading algorithms, the input data format of Recognition with Recurrent Neural Network model is defined, for outer foreign currency market
Different disks (Asia disk, European disk, America disk) building Trading Model of field, and for non-agricultural period and interest rate resolution period etc.
These can seriously affect the major event node of sentiment undertone, carry out the scientific transaction that places an order.Trading algorithms are used to market
Dynamic analysis, such as shake market, unilateral market under the different tendencies of market and all carry out the transaction that places an order, accomplish that day part is each
The accurate transaction of market.
Increase the real-time acquisition to market in Trading Model simultaneously.For monitoring, real-time is particularly important.
But thousands of broker's platforms are disposed simultaneously, to accomplish it is all can monitor in real time, technical difficulty quite it
Greatly.
To solve the above-mentioned problems, broker's platform lower for some market frequencies, we use efficient EA foot
This language can accomplish Millisecond delay acquisition.
Use significant advantage of the EA script in test process:
Hardware environment is identical.Each broker MT4 platform is taken under the computer environment under same configuration and identical network
It builds, all platforms locate same starting line.
Exchange hour is identical.For the order of synchronization, analysis evaluation and test is more objective.
Transaction more standardizes, and quantity on order is almost the same.EA transaction is different from such no chapter at random that manually places an order, all platforms
It is identical that quantity on order reaches 98%.(therein 2% is not both because some platforms will appear the case where going offline or disconnect automatically
Lead to not place an order)
It monitors quotations in real time, any data will not be missed.It ensure that the integrality of monitoring data.
For frequent broker's platform of offering, the present invention provides a kind of cloud computings to handle acquisition scheme, and use is multi-thread
Journey dual-mode network model, continues to optimize improvement on this basis, so that market monitoring error time is no more than 10 milliseconds, really
Realize the seamless monitoring of annual all the period of time to market.
Step 4: data import Cloud Server.Mass data storage realizes there are strict requirements to technology, needs to accomplish
Each pen all precisely monitors.These data have very great meaning for history verification and history reproduction etc..In terms of market,
For the daily quotation quantity of average every broker at 60,000, then 1,000 one day data volumes are exactly 6,000 ten thousand, add up the number that gets off
It is very big according to amount.We have carried out comprehensive record to each data, when the local zone time comprising this data, server
Between, buying price, selling price, newest valence, the whole quotation informations of point difference etc., it is in addition automatic again to generate 1 minute, 5 minutes, 30 minutes, 1 in real time
The multicycles data such as hour, 4 hours, 1 day, 1 week, this cycle data are used for the detection processing in later period.This calculates the processing of data
Method requires extremely stringent.
In terms of transaction, the present invention is traded by sequencing, realizes that machine completion automated intelligent places an order, will be in MT4
Data are transmitted to the long short-term memory layer of Recognition with Recurrent Neural Network model, carry out the training of neural network model, pass through the instruction of system
Practice the parameter in method and adaptive estimation transaction, screens judge that special market (larger fluctuation occur or frequency is accelerated in real time
Deng), trading range covers all constant time ranges, guarantees the integrality and high efficiency of evaluation and test data as far as possible.By each order
Local zone time, server time, type, lower single quantity, lower single direction, lower single price, allow sliding points, stop loss valence, be only full of valence,
Server is closed a position time, local close a position time, settlement price, quantity of closing a position, all information such as sliding point of closing a position all storages in real time.
And all sheet information, it is whole comprising all sheet types such as BUYLIMIT, SELLLIMIT, BUYSTOP, SELLSTOP
Storage, establishes huge Database Systems, designs rigorous data dictionary, formulates specific data number, real-time tracking data road
Diameter, each family all strict classifications, is truly realized data mostly without unrest.
Step 5: for the monitoring of market, market abnormal determination mark that the neural network model studied according to the present invention obtains
Standard is divided into market and interrupts (data outage or server interrupt), quotations abnormal (market price deviates that mean value is larger or quotation
Mistake) situations such as, by diagnostic result real-time Transmission to alarm center, notice is manually handled or is intervened, and is avoided tighter
The consequence of weight.
The present invention has done more careful design for alarm center, can receive monitoring data information in real time, according to connecing
The market coding and timestamp, intelligent Matching abnormal conditions received, and addressee is notified in a manner of mail or short message etc. in time,
To achieve the purpose that handle in time.
Alarm center is provided with multiple Data Nodes, we can carry out flexible configuration to it, can be sent to simultaneously more
A addressee, and different contents of announcement can also be set for different addressees, have safely controllable and no-delay
Feature.It can be directed to the different numerical value of different parameter settings in alarm center, and can voluntarily optimize, realize and do intelligence
Change function, is greatly saved manpower and material resources.
Step 6: in terms of market other than real time monitoring, the present invention continues to optimize supposition according to multiple historical datas, most
A set of scientific and reasonable marking scheme is realized eventually.
Its main contents includes:
Quotations number and frequency (maximum times, minimum number；Maximum frequency, minimum frequency)；
Server breaks number and server disconnection reconnecting time, maximum duration and shortest time including disconnection reconnecting；
Point difference situation in one day of trade and in one section of period (it is poor to specifically include maximum point difference, equalization point difference, point
Point difference between distribution situation, 4:30- morning, 6:30 morning expands situation；)；
More lists interest overnight of the daily each kind of statistics, empty single interest overnight, and thus calculate its interest and.
Above each data during distinguishing non-agricultural market period and interest rate resolution market period.
According to these data, the present invention obtains every exponential number by calculating variance yields respectively.
If in one group of quotation number data x1, x2, x3 ... xn, the average of each group of data and theyDifference square
It is respectivelySo we are with their average
To measure the fluctuation size of this group of data, formula are as follows:
(the wherein average value that x is this group of data).
Similarly, in quotation frequency, broken string number, server reconnection time, point difference maximum value, the poor minimum value of point, the poor expansion of point
Greatly, more simple interests, empty simple interest, more empty interests and these wholes carry out depth calculation again, finally obtain market environmental index
The poor environmental index (Spread environment Index) of (Market environment Index), point, cost overnight refer to
Number (Swap environment Index).Wherein the index value full marks of each single item are 10 points, pass through these calculated numerical value
The superiority and inferiority of the platform market environment can intuitively be told.
Step 7: being the analysis to market environment above, wherein a most important ring is evaluation and test to trading environment.This
It is the aspect that the present invention studies emphatically, because the superiority and inferiority of trading environment is the most important thing for a user.Previous continuous
Research test in it was found that some differences platform, during customer transaction, can automatically modification user lower unit price
Lattice generate huge sliding point with this, and most of profit of user has been stolen in complete secrecy.(sliding point refers to the point to place an order
Position and the point that finally strikes a bargain have gap, as difference and commission, sliding point be the cost of transaction always.Valence in real trade
The sliding point of lattice forward direction and the sliding point of negative sense are likely to occur.) it there are also some cases is bad broker in Background control user
Transaction data, that voluntarily modifies user opens settlement price, causes the great loss of user.Therefore the monitoring of trading environment is carried out, it is
The most important thing of the invention, equally technically there is also very big difficulty in face.For these realistic problems encountered, originally
Invention uses the following scheme to realize the real time monitoring to trading environment, and achieves extraordinary effect.
It is sufficiently large in order to make to evaluate and test sample, it is more representative, it is necessary first to have enough transaction data.If by hand
It places an order, inevitably will appear under mistake, the problems such as leaking down in actual conditions, and lower single opportunity of trade order is also easy by artificial
The influence of factor can not accomplish the time consistency of all platforms above exchange hour.Therefore the present invention is in trade order
Generate aspect and mainly use EA transaction system, formulated a set of effective Trading Model, trading session cover Asia disk,
All periods such as European disk, America disk, and the transaction that places an order also individually is carried out in non-agricultural period and interest rate resolution period.Under wherein
Folk prescription formula also contemplated all modes used in transaction, include market order, Limit Order, sheet etc..In the order of conclusion of the business, together
Having Self Trimming, being only full of and stop loss the modes such as close a position for sample setting, is also used in scalping under battle conditions with regard to all covering user in this way
All Activity mode.The present invention emphatically analyzes the progress subversiveness of these trade order data comprehensively and carries out careful analysis.
Each transaction data of user can all be recorded in log and save in a computer in MT4 software, these numbers
According to including the local zone time of order, open settlement price and order direction.The present invention starts with from the transaction log of software first,
By computer language to some crucial word analysis in log, each data in log, intelligent discrimination are efficiently parsed one by one
The data of each order are simultaneously put in storage in real time, all information datas comprising trade order.
On the other hand, in order to make up some loss of data in special circumstances in transaction log (under such as some platforms
The information such as manual order peace warehouse receipt fail to automatically save log), the ad hoc a set of order tracking technique for counting out oneself of the present invention,
In conjunction with EA tool, for single peaceful warehouse receipt that opens a position, then the newest valence of current cultivars when automatically recording lower single is remembered in real time again
Final knock-down price and corresponding timestamp after trading successfully under record, sit in the right seat and are sequentially recorded in database.
The present invention mainly passes through both schemes and works along both lines, and is complementary to one another, and records each transaction data, then root
According to these data depth analysis again, thus transactions velocity, the positive sliding point of transaction, the negative sliding point of transaction etc. to calculate each order
Data information.
The present invention focuses on the data information studied: transaction most jogging speed, transaction prestissimo, transaction average speed；Order
The negative sliding point of positive sliding point, order and the sliding point probability of generation derived, then calculate by pen according to these data, generate final
Trading environment index (Trading environment Index).
Wherein the calculating of trading environment index is calculated according to covariance and is realized:
Example: the calculating (X1, X2 ... ... indicate the average speed of each platform) to speed
If X=(X1, X2..., XN)T
For n n-dimensional random variable n, claim matrix
For n n-dimensional random variable n
The covariance matrix (covariance matrix) of speed,
cij=Cov (Xi, Xj), i, j=1,2 ..., n are the component X of XiAnd XjCovariance
Wherein
c11=E [X1-E(X1)]2, c12=E [X1-E(X1)][X2-ε(X2)]
c21=E [X2-E(X2)][X1-E(X1)], c22=E [X2-E(X2)]2
Step 8: the four big indexes by obtaining Platform Analysis with upper type: putting poor environmental index, market environmental index, overnight
Cost index, trading environment index.The present invention carries out a comprehensive scoring analysis in conjunction with this four exponent datas.According to
Different weights is arranged in influence situation of each single item to transaction in practice, finally obtains a composite index.
The last present invention again thoroughly integrates these data, generates day index, all indexes and monthly index respectively, passes through
Check this composite index, user can intuitively tell the superiority and inferiority of each broker's platform.
The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention.The technology of the industry
Personnel are it should be appreciated that the present invention is not limited to the above embodiments, and the above embodiments and description only describe this
The principle of invention, without departing from the spirit and scope of the present invention, various changes and improvements may be made to the invention, these changes
Change and improvement all fall within the protetion scope of the claimed invention.The claimed scope of the invention by appended claims and its
Equivalent thereof.
Claims (10)
1. a kind of Data Detection and analysis method for financial intermediary MT4 foreign exchange transaction software, which is characterized in that the party
Method specifically includes the following steps:
S1: the financial broker more than any one family to be monitored progress firm offer is opened an account and is injected big at home using internet
True fund is measured, all Account Types opened an account all are unified standard account；
S2: to ensure to monitor the just of environment, the MT4 software creation for all financial broker to be monitored is unified and high
The computer techno-stress operation hardware platform matched；
S3: writing trading algorithms, the input data format of Recognition with Recurrent Neural Network model is defined, for the different disk structures of foreign exchange market
It builds Trading Model, and resolves the period for non-agricultural period and interest rate, real-time monitoring is carried out to market, when monitoring that market are violent
Uniformly placed an order transaction when fluctuation；
S4: data import Cloud Server, mass data storage, in terms of market, to each number of each monitoring object
According to being recorded comprehensively, local zone time, server time, buying price, selling price, newest valence comprising this data, the whole rows of point difference
Feelings information, it is in addition automatic again to generate 1 minute, 5 minutes, 30 minutes, 1 hour, 4 hours, 1 day, 1 week these multicycle number in real time
According to this cycle data is used for the detection processing in later period；It in terms of transaction, is traded by sequencing, realizes that machine completion is automatic
Intelligence places an order, and the data in MT4 are transmitted to the long short-term memory layer of Recognition with Recurrent Neural Network model, carry out neural network model
Training, by the parameter in training method and the adaptive estimation transaction of system, screening judgement in real time includes common market and height
Frequency market these market dynamic, trading range cover all constant time ranges, guarantee the integrality and high efficiency of evaluation and test data；It will be each
The local zone time of order, type, lower single quantity, lower single direction, lower single price, allows sliding points, stops loss server time
Valence, close a position time, local close a position time, settlement price, quantity of closing a position, the sliding all information of point of closing a position of valence, server that are only full of are whole
Storage in real time；And all sheet information, it include these all extensions of BUYLIMIT, SELLLIMIT, BUYSTOP, SELLSTOP
Single type is all put in storage, and establishes Database Systems, and design data dictionary is formulated specific data and numbered, real-time tracking data path,
Each family all strict classifications, make data mostly without unrest；
S5: for the monitoring of market, the market abnormality judgment criterions obtained according to the neural network model that step S3 is created, point
It is interrupted for market, the abnormal two kinds of situations of quotations, by diagnostic result real-time Transmission to alarm center, notice is manually handled
Or intervene；
S6: binding site difference environmental index, market environmental index, cost index overnight, trading environment exponent data carry out synthesis
Scoring analysis, the influence situation according to each single item in practice to transaction are arranged different weights, finally obtain a composite index,
Historical data finally is gathered according to the difference in period to these data again, generates day index, all indexes and monthly index respectively, leads to
It crosses and checks this composite index, user can intuitively tell the superiority and inferiority of each broker's platform.
2. the Data Detection and analysis method according to claim 1 for financial intermediary MT4 foreign exchange transaction software,
It is characterized in that, trading algorithms use the dynamic analysis to market in the step S3, as shaken under the different tendencies of market
Market, unilateral market all carry out the transaction that places an order, and accomplish the accurate transaction of each market of day part.
3. the Data Detection and analysis method according to claim 1 for financial intermediary MT4 foreign exchange transaction software,
It is characterized in that, the Trading Model in the step S3 can obtain in real time market.
4. the Data Detection and analysis method according to claim 3 for financial intermediary MT4 foreign exchange transaction software,
It is characterized in that, the market for the lower broker's platform of some market frequencies are obtained in real time using efficient EA script
Language can accomplish Millisecond delay acquisition.
5. the Data Detection and analysis method according to claim 3 for financial intermediary MT4 foreign exchange transaction software,
It is characterized in that, described use multithreading dual-mode network model for frequent broker's platform of offering, market is made to monitor error
Time is no more than 10 milliseconds, to realize the seamless monitoring of annual all the period of time to market.
6. the Data Detection and analysis method according to claim 1 for financial intermediary MT4 foreign exchange transaction software,
It is characterized in that, the alarm center can receive monitoring data information in real time, according to the market coding and timestamp received,
Intelligent Matching abnormal conditions, and addressee is notified with mail or short message mode in time, to handle in time.
7. the Data Detection and analysis method according to claim 1 for financial intermediary MT4 foreign exchange transaction software,
It is characterized in that, the alarm center is provided with multiple Data Nodes, flexible configuration can be carried out to it, can be sent to simultaneously
Multiple addressees, and different contents of announcement can also be set for different addressees, have safely controllable and no-delay
The characteristics of.
8. the Data Detection and analysis method according to claim 1 for financial intermediary MT4 foreign exchange transaction software,
It is characterized in that, described be directed in method to the real time monitoring in terms of market, it is provided with marking scheme, the program is to quotation frequency
Rate, broken string number, the server reconnection time, put poor maximum value, put poor minimum value, put poor expansions, more simple interests, sky simple interest,
More sky interests and these data obtain every exponential number by calculating variance yields respectively, then complete full depth and calculate,
It finally obtains market environmental index, put poor environmental index, cost index overnight；Wherein the index value full marks of each single item are 10 points,
The superiority and inferiority of the platform market environment can be intuitively told by these calculated numerical value.
9. the Data Detection and analysis method according to claim 1 for financial intermediary MT4 foreign exchange transaction software,
It is characterized in that, to transaction most jogging speed, transaction prestissimo, transaction average speed in the method；Order just sliding point, order
Negative sliding point and sliding point these data of probability of generation derived, calculate by pen according to covariance, generate final transaction ring
Border index.
10. the Data Detection according to claim 1 for financial intermediary MT4 foreign exchange transaction software and analysis side
Method, which is characterized in that the computer techno-stress operation hardware platform is configured to CPU: eight core 2.5GHz, memory: 32GB, band
It is wide: 100M.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910706060.8A CN110428333A (en) | 2019-08-01 | 2019-08-01 | For the Data Detection and analysis method of financial intermediary MT4 foreign exchange transaction software |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910706060.8A CN110428333A (en) | 2019-08-01 | 2019-08-01 | For the Data Detection and analysis method of financial intermediary MT4 foreign exchange transaction software |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110428333A true CN110428333A (en) | 2019-11-08 |
Family
ID=68413726
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910706060.8A Pending CN110428333A (en) | 2019-08-01 | 2019-08-01 | For the Data Detection and analysis method of financial intermediary MT4 foreign exchange transaction software |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110428333A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113222602A (en) * | 2021-05-19 | 2021-08-06 | 中国银行股份有限公司 | Foreign exchange transaction processing system and method of 5G message based on block chain |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5761386A (en) * | 1996-04-05 | 1998-06-02 | Nec Research Institute, Inc. | Method and apparatus for foreign exchange rate time series prediction and classification |
| CN101710306A (en) * | 2009-12-15 | 2010-05-19 | 中国科学院软件研究所 | Method and system for detecting software reliability |
| CN101794430A (en) * | 2010-02-11 | 2010-08-04 | 帅亿 | Fully-automatic stock operation server |
| CN107885999A (en) * | 2017-11-08 | 2018-04-06 | 华中科技大学 | A kind of leak detection method and system based on deep learning |
| CN108647518A (en) * | 2018-03-16 | 2018-10-12 | 广东工业大学 | A kind of Android platform malware detection method based on deep learning |
| CN108694480A (en) * | 2018-07-13 | 2018-10-23 | 西南石油大学 | Finance data prediction technique based on improved length memory network in short-term |
| CN109242207A (en) * | 2018-10-10 | 2019-01-18 | 中山大学 | A kind of Financial Time Series prediction technique based on deeply study |
| CN109461067A (en) * | 2018-09-12 | 2019-03-12 | 阿里巴巴集团控股有限公司 | A kind of detection method of foreign exchange quotation abnormal data, apparatus and system |
| CN109919363A (en) * | 2019-02-18 | 2019-06-21 | 上海金大师网络科技有限公司 | The price expectation method, device and equipment of financial product |
| CN109992502A (en) * | 2017-12-30 | 2019-07-09 | 中国移动通信集团四川有限公司 | Application software performance automatic detection analysis method, apparatus, equipment and medium |
-
2019
- 2019-08-01 CN CN201910706060.8A patent/CN110428333A/en active Pending
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5761386A (en) * | 1996-04-05 | 1998-06-02 | Nec Research Institute, Inc. | Method and apparatus for foreign exchange rate time series prediction and classification |
| CN101710306A (en) * | 2009-12-15 | 2010-05-19 | 中国科学院软件研究所 | Method and system for detecting software reliability |
| CN101794430A (en) * | 2010-02-11 | 2010-08-04 | 帅亿 | Fully-automatic stock operation server |
| CN107885999A (en) * | 2017-11-08 | 2018-04-06 | 华中科技大学 | A kind of leak detection method and system based on deep learning |
| CN109992502A (en) * | 2017-12-30 | 2019-07-09 | 中国移动通信集团四川有限公司 | Application software performance automatic detection analysis method, apparatus, equipment and medium |
| CN108647518A (en) * | 2018-03-16 | 2018-10-12 | 广东工业大学 | A kind of Android platform malware detection method based on deep learning |
| CN108694480A (en) * | 2018-07-13 | 2018-10-23 | 西南石油大学 | Finance data prediction technique based on improved length memory network in short-term |
| CN109461067A (en) * | 2018-09-12 | 2019-03-12 | 阿里巴巴集团控股有限公司 | A kind of detection method of foreign exchange quotation abnormal data, apparatus and system |
| CN109242207A (en) * | 2018-10-10 | 2019-01-18 | 中山大学 | A kind of Financial Time Series prediction technique based on deeply study |
| CN109919363A (en) * | 2019-02-18 | 2019-06-21 | 上海金大师网络科技有限公司 | The price expectation method, device and equipment of financial product |
Non-Patent Citations (1)
| Title |
|---|
| 姚小强 等: "基于树结构LSTM神经网络的金融时间序列预测", 《计算机应用》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113222602A (en) * | 2021-05-19 | 2021-08-06 | 中国银行股份有限公司 | Foreign exchange transaction processing system and method of 5G message based on block chain |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2021232588A1 (en) | Food safety risk assessment method, apparatus, device, and storage medium | |
| US10552735B1 (en) | Applied artificial intelligence technology for processing trade data to detect patterns indicative of potential trade spoofing | |
| CN110417721A (en) | Safety risk estimating method, device, equipment and computer readable storage medium | |
| CN106952159A (en) | A kind of real security risk control method, system and storage medium | |
| CN108053087A (en) | Anti money washing monitoring method, equipment and computer readable storage medium | |
| JP2004506959A (en) | Evaluation value prediction model when information is missing | |
| CN103678659A (en) | E-commerce website cheat user identification method and system based on random forest algorithm | |
| US20090055245A1 (en) | Survey fraud detection system and method | |
| JP2004500641A (en) | Method and system for automatically estimating credit score evaluation value | |
| KR101886943B1 (en) | A method for establishing power supply and demand early warning model based on dynamic clustering | |
| JP2004500642A (en) | Methods and systems for assessing cash flow recovery and risk | |
| CN110162445A (en) | The host health assessment method and device of Intrusion Detection based on host log and performance indicator | |
| CN112700324A (en) | User loan default prediction method based on combination of Catboost and restricted Boltzmann machine | |
| CN110262919A (en) | Abnormal data analysis method, device, equipment and computer readable storage medium | |
| Zhu et al. | [Retracted] Fuzzy Decision‐Making Analysis of Quantitative Stock Selection in VR Industry Based on Random Forest Model | |
| He et al. | A data-analytics approach for risk evaluation in peer-to-peer lending platforms | |
| Cao et al. | Detecting wash trade in the financial market | |
| CN114819494A (en) | Enterprise risk early warning method, device, equipment and medium | |
| CN110428333A (en) | For the Data Detection and analysis method of financial intermediary MT4 foreign exchange transaction software | |
| Li et al. | Stock price prediction based on generative adversarial network | |
| KR20050012699A (en) | Method on stock trading management | |
| CN117078266A (en) | Transaction data security detection method, device, computer equipment and storage medium | |
| Yunchao et al. | Research on the core competitiveness of pharmaceutical listed companies based on fuzzy comprehensive evaluation | |
| Wang | Early warning and monitoring analysis of financial accounting indicators of listed companies based on big data | |
| Sharma et al. | Prediction of Real-Time Estate Pricing using Train-Test Splitting Techniques |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20191108 |