CN109472399A - Consider the air ticket purchase decision method and system of uncertainty in traffic - Google Patents
Consider the air ticket purchase decision method and system of uncertainty in traffic Download PDFInfo
- Publication number
- CN109472399A CN109472399A CN201811238947.0A CN201811238947A CN109472399A CN 109472399 A CN109472399 A CN 109472399A CN 201811238947 A CN201811238947 A CN 201811238947A CN 109472399 A CN109472399 A CN 109472399A
- Authority
- CN
- China
- Prior art keywords
- date
- price
- indicate
- target flight
- air ticket
- 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
- 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
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0283—Price estimation or determination
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Development Economics (AREA)
- Theoretical Computer Science (AREA)
- Strategic Management (AREA)
- Data Mining & Analysis (AREA)
- Economics (AREA)
- Mathematical Analysis (AREA)
- Entrepreneurship & Innovation (AREA)
- General Business, Economics & Management (AREA)
- Game Theory and Decision Science (AREA)
- Finance (AREA)
- Human Resources & Organizations (AREA)
- Operations Research (AREA)
- Marketing (AREA)
- Mathematical Physics (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Optimization (AREA)
- Accounting & Taxation (AREA)
- Computational Mathematics (AREA)
- Quality & Reliability (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Probability & Statistics with Applications (AREA)
- Life Sciences & Earth Sciences (AREA)
- Algebra (AREA)
- Tourism & Hospitality (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)
Abstract
The present invention provides a kind of air ticket purchase decision method and systems for considering uncertainty in traffic, determine the target flight of purchase, history air ticket observation until obtaining the departure date to current date of target flight generates original data set based on history air ticket observation；Based on original data set, characteristic data set is generated；The multiple linear regression model of the corresponding different prediction steps of training；The Estimating Confidence Interval for calculating different prediction steps obtains current date to forecast price probability distribution daily between the departure date of target flight；Using the multiple sampling of forecast price probability distribution, a plurality of simulated price path is obtained, a plurality of simulated price path is retrodicted, obtains purchase Proposals.By the Estimating Confidence Interval of multiple linear regression, calculates the expectation of the following cost and make a policy, make up the theoretical vacancy in current air ticket purchase decision field to a certain extent.
Description
Technical field
The present invention relates to purchase consumption decision fields, and in particular, to a kind of air ticket purchase for considering uncertainty in traffic
Decision-making technique and system.
Background technique
With the gradually development of aircraft industry, its income improves using complicated pricing strategy in most of airline, from
And the admission fee of aircraft is caused to show high uncertainty, and since factors based on price are not public institute
Know, ordinary consumer is often difficult to hold the alteration trend of price.Existing method mostly from the angle of airline, with
Yield management strategy is designed for the purpose of raising corporate income.A kind of purchase plan is designed from the angle for saving expenditure for buyer
Slightly, there is significant application value for the majority of consumers and online ticket service company.
According to common sense, a possibility that more early purchase obtains lower price, is higher, but according to the system to real machine ticket data
Meter, distribution of the lowest price before the date, which levels off to, to be uniformly distributed.In existing research, a kind of method is when utilizing
Between sequence prediction method forecast price and threshold value be set carry out decision, but these methods cannot well adapt to air ticket
The dynamic of list price and the trend that cannot completely hold price change in the entire time.Air ticket purchase is that a kind of timing decision is asked
Topic, the mode of optimum stopping theory probabilistic model, which provides optimality, to be proved, but the theory has strong assumption condition, i.e., random to become
Amount is obeyed with distribution, this condition limits its universality.
Ticket price prediction refers to the specific at following one or more time points of some given ticket of prediction
Price.It accurately predicts ticket price, no matter budget is saved for consumer, or be that air ticket dealer holds market trends
There is greatly positive facilitation.Ticket price prediction is a typical recurrence learning problem, and the admission fee as output valve is
The variable of one continuous type.In current research and field is practiced, it is less directly against the theory and method of ticket price prediction.
For time series forecasting problem, more common method is multiple linear regression.Since its model is simple and can be adapted for more
The characteristics of metavariable, multiple linear regression have been applied at present in many practical problems.Due to the pricing strategy of airline
The error of the information dissymmetry of dynamic change and both parties, ticket price prediction is inevitable.And it uses wrong
It is bad that the prediction result of difference as the basis of purchase decision may result in tactful effect.It therefore, should when doing decision
Consider the accuracy of prediction result.The Estimating Confidence Interval of multiple linear regression is to provide mean of dependent variable to given independent variable
Interval estimation the uncertainty of prediction result can be measured by the estimation.
Summary of the invention
For the defects in the prior art, the object of the present invention is to provide a kind of air ticket purchases for considering uncertainty in traffic
Decision-making technique and system.
A kind of air ticket purchase decision method of the consideration uncertainty in traffic provided according to the present invention, comprising:
It obtains original prices data step: determining the target flight of purchase, obtain the departure date of target flight to currently
History air ticket observation until date generates original data set based on history air ticket observation；
Training dataset step: being based on original data set, generates characteristic data set；
Multi-step prediction step: the multiple linear regression model of the corresponding different prediction steps of characteristic data set training is utilized；
Estimating Confidence Interval step: calculating the Estimating Confidence Interval of different prediction steps, obtains current date to target and navigates
Daily forecast price probability distribution between the departure date of class；
It generates purchase strategy step: using the multiple sampling of forecast price probability distribution, obtaining a plurality of simulated price path,
A plurality of simulated price path is retrodicted, minimum cost desired value, current date cost are obtained, if minimum cost desired value is small
It in current date cost, then buys Proposals and postpones purchase, otherwise, then suggest that current date is bought.
Preferably, the characteristic data set mainly includes that target flight closes on date price before current dateTarget flight closes on date price average before current dateTarget flight is being worked as
Date price minimum value is closed on before the preceding dateTarget flight closes on the flight number that the date sets out and closes on the date
PriceNumber of days of the target flight apart from departure date, target flight setting time before current date
Any one of interior price raising number is appointed multinomial；
Wherein,To indicate that target pattern r departure date is i days before d takes off admission fees；Subscript i indicates number of days；R table
Show target flight；The departure date of d expression target flight；Indicate that target pattern r departure date is i days before d takes off tickets
Valence summation.
Preferably, the current date to forecast price probability distribution daily between the departure date of target flight indicates
Are as follows:
In formula, P indicates forecast price probability distribution；
The departure date of T expression target flight；
S indicates current date；
Indicate that current date is general to forecast price daily between the departure date of target flight
Rate distribution；
hsIndicate historical price data；
F(hs) indicate price expectation model.
Preferably, the multiple linear regression model indicates are as follows:
In formula, piIndicate i-th of price expectation value；
Indicate multi-C vector；
eiIndicate the random error of i-th of price expectation value；
Indicate the transposition of i-th of vector.
Preferably, in the characteristic data set training, Least Square Method parameter:
βOLS=X (XTX)-1XTP,
Estimated value are as follows:
Residual sum of squares (RSS) are as follows:
Mean square error are as follows:
MSE=SSE/ (n-p)
Estimating Confidence Interval are as follows:
In formula,It can be replaced with normal distribution；
And then obtain the forecast price probability distribution on t date are as follows:
In formula, βOLSIndicate least-squares estimation parameter；
X indicates eigenmatrix；
XTIndicate the transposition of eigenmatrix；
P indicates price vector；
Indicate price expectation value；
E (P | x) indicates the expectation of predicted value；
xTIndicate the transposition of vector；
β expression parameter vector；
SSE is indicated and variance；
Indicate remainder square；
MSE indicates mean square deviation；
N indicates sample number；
P indicates intrinsic dimensionality；
Indicate estimator t distribution；
Indicate confidence interval；
Indicate the transposition of feature vector；
xfIndicate feature vector；
Indicate the distribution of t date forecast price；
CI () indicates Estimating Confidence Interval；
Indicate t date prediction probability price；
Indicate t date forecast price confidence interval.
A kind of air ticket purchase decision system of the consideration uncertainty in traffic provided according to the present invention, comprising:
It obtains original prices data module: determining the target flight of purchase, obtain the departure date of target flight to currently
History air ticket observation until date generates original data set based on history air ticket observation；
Training dataset module: being based on original data set, generates characteristic data set；
Multi-step prediction module: the multiple linear regression model of the corresponding different prediction steps of characteristic data set training is utilized；
Estimating Confidence Interval module: calculating the Estimating Confidence Interval of different prediction steps, obtains current date to target and navigates
Daily forecast price probability distribution between the departure date of class；
It generates purchase policy module: using the multiple sampling of forecast price probability distribution, obtaining a plurality of simulated price path,
A plurality of simulated price path is retrodicted, minimum cost desired value, current date cost are obtained, if minimum cost desired value is small
It in current date cost, then buys Proposals and postpones purchase, otherwise, then suggest that current date is bought.
Compared with prior art, the present invention have it is following the utility model has the advantages that
The present invention provides consumer's Cost Model of air ticket purchase, considers the uncertainty of ticket price prediction, decision base
To the complete analysis of the complete price route before departure date when decision.By Estimating Confidence Interval price distribution and use
Multiple sampling Simulation and Decision process can fully consider bring effect after postponing purchase.
Detailed description of the invention
Upon reading the detailed description of non-limiting embodiments with reference to the following drawings, other feature of the invention,
Objects and advantages will become more apparent upon:
Fig. 1 is air ticket purchase decision flow chart.
Specific embodiment
The present invention is described in detail combined with specific embodiments below.Following embodiment will be helpful to the technology of this field
Personnel further understand the present invention, but the invention is not limited in any way.It should be pointed out that the ordinary skill of this field
For personnel, without departing from the inventive concept of the premise, several changes and improvements can also be made.These belong to the present invention
Protection scope.
By the Estimating Confidence Interval of multiple linear regression, the uncertainty of ticket price prediction result is analyzed, in this base
The expectation of the following cost is calculated on plinth by multiple repairing weld Simulation and Decision process, and is made a policy on this basis, with certain journey
The theoretical vacancy in current air ticket purchase decision field is made up on degree.
A kind of air ticket purchase decision method of the consideration uncertainty in traffic provided according to the present invention, comprising:
It obtains original prices data step: determining the target flight of purchase, obtain the departure date of target flight to currently
History air ticket observation until date generates original data set based on history air ticket observation；
Training dataset step: being based on original data set, generates characteristic data set；
Multi-step prediction step: the multiple linear regression model of the corresponding different prediction steps of characteristic data set training is utilized；
Estimating Confidence Interval step: calculating the Estimating Confidence Interval of different prediction steps, obtains current date to target and navigates
Daily forecast price probability distribution between the departure date of class；
It generates purchase strategy step: using the multiple sampling of forecast price probability distribution, obtaining a plurality of simulated price path,
Based on minimum expectation cost algorithms, a plurality of simulated price path is retrodicted, obtains minimum cost desired value, current date generation
Valence buys Proposals and postpones purchase if minimum cost desired value is less than current date cost, otherwise, then suggests working as the day before yesterday
Phase purchase.
A kind of air ticket purchase decision system of the consideration uncertainty in traffic provided according to the present invention, comprising:
It obtains original prices data module: determining the target flight of purchase, obtain the departure date of target flight to currently
History air ticket observation until date generates original data set based on history air ticket observation；
Training dataset module: being based on original data set, generates characteristic data set；
Multi-step prediction module: the multiple linear regression model of the corresponding different prediction steps of characteristic data set training is utilized；
Estimating Confidence Interval module: calculating the Estimating Confidence Interval of different prediction steps, obtains current date to target and navigates
Daily forecast price probability distribution between the departure date of class；
It generates purchase policy module: using the multiple sampling of forecast price probability distribution, obtaining a plurality of simulated price path,
Based on minimum expectation cost algorithms, a plurality of simulated price path is retrodicted, obtains minimum cost desired value, current date generation
Valence buys Proposals and postpones purchase if minimum cost desired value is less than current date cost, otherwise, then suggests working as the day before yesterday
Phase purchase.
The air ticket purchase decision system provided by the invention for considering uncertainty in traffic, can be by considering that prediction is uncertain
Property air ticket purchase decision method step process realize.Those skilled in the art can will consider the air ticket of uncertainty in traffic
Purchase decision method is interpreted as the preference of the air ticket purchase decision system for considering uncertainty in traffic.
Specifically, the characteristic data set mainly includes that target flight closes on date price before current dateTarget flight closes on date price average before current dateTarget flight is being worked as
Date price minimum value is closed on before the preceding dateTarget flight closes on the flight number that the date sets out and closes on the date
PriceNumber of days of the target flight apart from departure date, target flight setting time before current date
Any one of interior price raising number is appointed multinomial；
Wherein,To indicate that target pattern r departure date is i days before d takes off admission fees；Subscript i indicates number of days；R table
Show target flight；The departure date of d expression target flight；Indicate that target pattern r departure date is i days before d takes off tickets
Valence summation.
Specifically, the current date to forecast price probability distribution daily between the departure date of target flight indicates
Are as follows:
In formula, P indicates forecast price probability distribution；
The departure date of T expression target flight；
S indicates current date；
Indicate that current date is general to forecast price daily between the departure date of target flight
Rate distribution；
hsIndicate historical price data；
F(hs) indicate price expectation model
Specifically, the multiple linear regression model indicates are as follows:
In formula, piIndicate i-th of forecast price；
β indicates multi-C vector；
ei, indicate the random error of i-th of price expectation value；
Indicate the transposition of i-th of vector.
Specifically, in the characteristic data set training, Least Square Method parameter:
βOLS=X (XTX)-1XTP,
Estimated value are as follows:
Residual sum of squares (RSS) are as follows:
Mean square error are as follows:
MSE=SSE/ (n-p)
Estimating Confidence Interval are as follows:
In formula,It can be replaced with normal distribution；
And then obtain the forecast price probability distribution on t date are as follows:
In formula, βOLSIndicate least-squares estimation parameter；
X indicates eigenmatrix；
XTIndicate the transposition of eigenmatrix；
P indicates price vector；
Indicate price expectation value；
E (P | x) indicates the expectation of predicted value；
xTIndicate the transposition of vector；
β expression parameter vector；
SSE is indicated and variance；
Indicate remainder square；
MSE indicates mean square deviation；
N indicates sample number；
P indicates intrinsic dimensionality；
Indicate estimator t distribution；
Indicate confidence interval；
Indicate the transposition of feature vector；
xfIndicate feature vector；
Indicate the distribution of t date forecast price；
CI () indicates Estimating Confidence Interval；
Indicate t date prediction probability price；
Indicate t moment forecast price confidence interval.
Core of the invention thought is ticket price sequence being regarded as time series, by multiple linear regression to setting out
Price before day carries out multi-step prediction, then utilizes the Estimating Confidence Interval method predictive metrics knot of multiple linear regression model
The uncertainty of fruit.The data that prediction technique is relied on, the predominantly historical price of air ticket observe data.
Assuming that target flight departure date is T, need to provide building for purchase or waiting at each moment before the T moment
View, decision process terminates if providing purchase and suggesting or reach departure date.Purchase Cost Model is carried out below brief
Explanation.When consumer purchases the air ticket, cabin or purchase failure wind are risen caused by purchase it should also be taken into account that postponing other than ticket price
Danger.Meeting is higher and higher as departure date closes on for the risk, and significantly mentions within shorter a period of time before day of setting out
It is high.Therefore the risk part bought in cost is indicated with exponential model.It is indicated to which the consumer of t moment is bought cost
Are as follows:
Cost (t, θ, δ)=pt+θδT-t,
Wherein θ is the greateset risk constant risk of T moment (corresponding), and δ is risk discount factor, and 0 δ≤1 <.
Method of the invention is made of two stages: forecast period and decision phase.In forecast period, according to historical price
Observation data predict the daily price distribution before current time to, indicate are as follows:
Wherein hsIt is historical price data.Direct strategy is used to the multi-step prediction problem in the present invention, i.e., to each step
Length establishes prediction model respectively.
Price distribution prediction technique is illustrated below.Basic forecast model of the invention is multiple linear regression mould
Type indicates are as follows:
It is wherein multi-C vector, ei,It is random error.The present invention uses Least Square Method parameter:
βOLS=X (XTX)-1XTP,
So estimated value table is shown as:
Residual sum of squares (RSS) are as follows:
Mean square error are as follows:
MSE=SSE/ (n-p).
Estimating Confidence Interval are as follows:
In this problem, since sample size is very big,It can be replaced with normal distribution.The price distribution of t moment is pre-
It surveys are as follows:
Second stage is the decision phase.The core of decision strategy is, more to future price according to price distribution prediction result
Secondary sampling, and Simulation and Decision process, if to calculate the expectation that the minimum cost that may be obtained after purchase is postponed in selection:
E[π(ps+1,δ,θ,P)]
The principle of decision is, if the following expectation minimum cost is lower than current time, selects to postpone and buy, otherwise select
It buys, indicates at current time are as follows:
E[π(ps+1,δ,θ,P)]≤(ps+θδT-s)
The present invention provides the algorithm of desired cost: according to price expectation as a result, available current time to day of setting out
Between daily price distributionBecause the uncertainty of forecast price, use is according to prediction after sampling
The available corresponding probability of the probability distribution of price, uses probability as Weight the result of multiple repairing weld, can be obtained
The expectation cost at the moment.Because the following minimum cost needs multiple recursive calculation, and the estimation of forecast price distribution is all only
Vertical, in order to improve efficiency, calculated using reverse method.It is simulated during retrodicting and the complete of decision is carried out using the decision strategy
It has suffered journey.Because forecast price is not single value but a series of distributions, a variety of different costs can be generated according to these distributions
Path, in order to fully consider uncertainty in traffic, the method that the present invention uses multiple repairing weld.The pseudo- generation of minimum cost Expectation Algorithm
Code is as shown in Table 1.
Preference of the invention is further elaborated below in conjunction with Fig. 1.
Embodiment of the present invention follows the steps below:
Step 1: determining the target flight of purchase, and all history air tickets observation until obtaining the flight to current date
Value.For example, target flight is set to Beijing on October 1st, 2018 to the MU5137 flight in Shanghai, current date is in September, 2018
10, then it should include all ticket price records observed in September in 2018 10 days and in the past in original data set.Make
WithTo indicate that course line r departure date is i days before d takes off admission fees.
Step 2: generating characteristic data set on the basis of original data set.Feature includes: the flight in the observational day
Date price is closed on before phaseThe flight closes on date price average before the observation dateThe flight closes on date price minimum value before the observation dateThe flight closes on the date and goes out
The flight number of hair closes on the price on dateApart from the number of days of departure date, the flight is before observing the date
Price raising number in a period of time.
Step 3: utilizing the multiple linear regression model of the corresponding different prediction steps of characteristic data set training.
Step 4: calculating the Estimating Confidence Interval of different prediction steps, obtain from 11 days to 2018 October 1 of September in 2018
Daily forecast price probability distribution between day.
Step 5: obtaining a plurality of simulated price path using forecast price distribution multiple sampling.It is calculated according to minimum expectation cost
Method, simulated price path was retrodicted forward since on October 1st, 2018, on each date current cost and the following minimum cost
Compare, is minimized.The process was until 11 end of day of September in 2018, if obtaining the selection on the 10th of September in 2018 postpones purchase most
Small cost expectation.If the desired value 10, costs of less than 2018 on Septembers, Proposals postpone purchase, otherwise in the day
Phase purchase.
The pseudo table of one minimum cost Expectation Algorithm of table
One skilled in the art will appreciate that in addition to realizing system provided by the invention in a manner of pure computer readable program code
It, completely can be by the way that method and step be carried out programming in logic come so that provided by the invention other than system, device and its modules
System, device and its modules are declined with logic gate, switch, specific integrated circuit, programmable logic controller (PLC) and insertion
The form of controller etc. realizes identical program.So system provided by the invention, device and its modules may be considered that
It is a kind of hardware component, and the knot that the module for realizing various programs for including in it can also be considered as in hardware component
Structure；It can also will be considered as realizing the module of various functions either the software program of implementation method can be Hardware Subdivision again
Structure in part.
Specific embodiments of the present invention are described above.It is to be appreciated that the invention is not limited to above-mentioned
Particular implementation, those skilled in the art can make a variety of changes or modify within the scope of the claims, this not shadow
Ring substantive content of the invention.In the absence of conflict, the feature in embodiments herein and embodiment can any phase
Mutually combination.
Claims (6)
1. a kind of air ticket purchase decision method for considering uncertainty in traffic characterized by comprising
It obtains original prices data step: determining the target flight of purchase, obtain the departure date of target flight to current date
Until history air ticket observation, based on history air ticket observation generate original data set；
Training dataset step: being based on original data set, generates characteristic data set；
Multi-step prediction step: the multiple linear regression model of the corresponding different prediction steps of characteristic data set training is utilized；
Estimating Confidence Interval step: calculating the Estimating Confidence Interval of different prediction steps, obtains current date to target flight
Daily forecast price probability distribution between departure date；
It generates purchase strategy step: using the multiple sampling of forecast price probability distribution, a plurality of simulated price path is obtained, to more
Simulated price path is retrodicted, and minimum cost desired value, current date cost are obtained, and is worked as if minimum cost desired value is less than
Preceding date cost then buys Proposals and postpones purchase, otherwise, then suggests that current date is bought.
2. the air ticket purchase decision method according to claim 1 for considering uncertainty in traffic, which is characterized in that the spy
Sign data set owner will include that target flight closes on date price before current dateTarget flight is being worked as
Date price average is closed on before the preceding dateTarget flight closes on date price before current date
Minimum valueTarget flight closes on the price that the flight number that the date sets out closes on the dateTarget boat
Headway period is from any one of price raising number before current date in setting time of number of days, target flight of departure date or appoints
It is multinomial；
Wherein,To indicate that target pattern r departure date is i days before d takes off admission fees；Subscript i indicates number of days；R indicates mesh
Mark flight；The departure date of d expression target flight；The admission fee for indicating that target pattern r departure date is i days before d takes off is total
With.
3. the air ticket purchase decision method according to claim 1 for considering uncertainty in traffic, which is characterized in that described to work as
Daily forecast price probability distribution is expressed as between preceding date to the departure date of target flight:
In formula, P indicates forecast price probability distribution；
The departure date of T expression target flight；
S indicates current date；
Indicate current date to forecast price probability daily between the departure date of target flight point
Cloth；
hsIndicate historical price data；
F(hs) indicate price expectation model.
4. the air ticket purchase decision method according to claim 1 for considering uncertainty in traffic, which is characterized in that described more
First linear regression model (LRM) indicates are as follows:
In formula, piIndicate i-th of price expectation value；
β indicates multi-C vector；
eiIndicate the random error of i-th of price expectation value；
Indicate the transposition of i-th of vector.
5. the air ticket purchase decision method according to claim 1 for considering uncertainty in traffic, which is characterized in that the spy
It levies in data set training, Least Square Method parameter:
βOLS=X (XTX)-1XTP,
Estimated value are as follows:
Residual sum of squares (RSS) are as follows:
Mean square error are as follows:
MSE=SSE/ (n-p)
Estimating Confidence Interval are as follows:
In formula,It can be replaced with normal distribution；
And then obtain the forecast price probability distribution on t date are as follows:
In formula, βOLSIndicate least-squares estimation parameter；
X indicates eigenmatrix；
XTIndicate the transposition of eigenmatrix；
P indicates price vector；
Indicate price expectation value；
E (P | x) indicates the expectation of predicted value；
xTIndicate the transposition of vector；
β expression parameter vector；
SSE is indicated and variance；
Indicate remainder square；
MSE indicates mean square deviation；
N indicates sample number；
P indicates intrinsic dimensionality；
Indicate estimator t distribution；
Indicate confidence interval；
Indicate the transposition of feature vector；
xfIndicate feature vector；
Indicate the distribution of t date forecast price；
CI () indicates Estimating Confidence Interval；
Indicate t date prediction probability price；
Indicate t date forecast price confidence interval.
6. a kind of air ticket purchase decision system for considering uncertainty in traffic characterized by comprising
It obtains original prices data module: determining the target flight of purchase, obtain the departure date of target flight to current date
Until history air ticket observation, based on history air ticket observation generate original data set；
Training dataset module: being based on original data set, generates characteristic data set；
Multi-step prediction module: the multiple linear regression model of the corresponding different prediction steps of characteristic data set training is utilized；
Estimating Confidence Interval module: calculating the Estimating Confidence Interval of different prediction steps, obtains current date to target flight
Daily forecast price probability distribution between departure date；
It generates purchase policy module: using the multiple sampling of forecast price probability distribution, a plurality of simulated price path is obtained, to more
Simulated price path is retrodicted, and minimum cost desired value, current date cost are obtained, and is worked as if minimum cost desired value is less than
Preceding date cost then buys Proposals and postpones purchase, otherwise, then suggests that current date is bought.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811238947.0A CN109472399A (en) | 2018-10-23 | 2018-10-23 | Consider the air ticket purchase decision method and system of uncertainty in traffic |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811238947.0A CN109472399A (en) | 2018-10-23 | 2018-10-23 | Consider the air ticket purchase decision method and system of uncertainty in traffic |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109472399A true CN109472399A (en) | 2019-03-15 |
Family
ID=65663978
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811238947.0A Pending CN109472399A (en) | 2018-10-23 | 2018-10-23 | Consider the air ticket purchase decision method and system of uncertainty in traffic |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109472399A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110766227A (en) * | 2019-10-24 | 2020-02-07 | 沈阳民航东北凯亚有限公司 | Fare prediction method and device |
| CN111028020A (en) * | 2019-12-17 | 2020-04-17 | 上海携程商务有限公司 | Yield management method, system, medium and electronic device for flight bag machine scene |
| CN112541629A (en) * | 2020-12-10 | 2021-03-23 | 携程旅游网络技术（上海）有限公司 | Prediction method, system, electronic device and storage medium for airport selection |
| CN115809903A (en) * | 2023-02-06 | 2023-03-17 | 北京四方启点科技有限公司 | Ticket buying method and device, electronic equipment and storage medium |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2004088476A2 (en) * | 2003-03-27 | 2004-10-14 | University Of Washington | Performing predictive pricing based on historical data |
| US8200514B1 (en) * | 2006-02-17 | 2012-06-12 | Farecast, Inc. | Travel-related prediction system |
| KR20140065917A (en) * | 2012-11-22 | 2014-05-30 | 정성원 | Auction-based reservation system and auction-based reservation method |
| CN105869017A (en) * | 2016-03-29 | 2016-08-17 | 上海携程商务有限公司 | Method and system for predicting ticket prices |
| CN105976039A (en) * | 2016-05-06 | 2016-09-28 | 上海交通大学 | Hybrid purchase decision method based on air ticket price predication |
| US20170061555A1 (en) * | 2015-08-24 | 2017-03-02 | Mastercard International Incorporated | Method and system for predicting lowest airline ticket fares |
| CN107274215A (en) * | 2017-05-26 | 2017-10-20 | 黑龙江大学 | Flight prices Forecasting Methodology, device, equipment and storage medium |
| US20170323526A1 (en) * | 2013-10-21 | 2017-11-09 | Intralot, Inc. | Multi-jurisdictional progressive fast play lottery games with pool replenishment and associated methods |
| US20170364837A1 (en) * | 2016-06-20 | 2017-12-21 | Mastercard International Incorporated | Methods and Apparatus for Predicting Dynamic Pricing |
| CN107506435A (en) * | 2017-08-23 | 2017-12-22 | 深圳市活力天汇科技股份有限公司 | A kind of budget fare querying method based on price expectation |
-
2018
- 2018-10-23 CN CN201811238947.0A patent/CN109472399A/en active Pending
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2004088476A2 (en) * | 2003-03-27 | 2004-10-14 | University Of Washington | Performing predictive pricing based on historical data |
| US8200514B1 (en) * | 2006-02-17 | 2012-06-12 | Farecast, Inc. | Travel-related prediction system |
| KR20140065917A (en) * | 2012-11-22 | 2014-05-30 | 정성원 | Auction-based reservation system and auction-based reservation method |
| US20170323526A1 (en) * | 2013-10-21 | 2017-11-09 | Intralot, Inc. | Multi-jurisdictional progressive fast play lottery games with pool replenishment and associated methods |
| US20170061555A1 (en) * | 2015-08-24 | 2017-03-02 | Mastercard International Incorporated | Method and system for predicting lowest airline ticket fares |
| CN105869017A (en) * | 2016-03-29 | 2016-08-17 | 上海携程商务有限公司 | Method and system for predicting ticket prices |
| CN105976039A (en) * | 2016-05-06 | 2016-09-28 | 上海交通大学 | Hybrid purchase decision method based on air ticket price predication |
| US20170364837A1 (en) * | 2016-06-20 | 2017-12-21 | Mastercard International Incorporated | Methods and Apparatus for Predicting Dynamic Pricing |
| CN107274215A (en) * | 2017-05-26 | 2017-10-20 | 黑龙江大学 | Flight prices Forecasting Methodology, device, equipment and storage medium |
| CN107506435A (en) * | 2017-08-23 | 2017-12-22 | 深圳市活力天汇科技股份有限公司 | A kind of budget fare querying method based on price expectation |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110766227A (en) * | 2019-10-24 | 2020-02-07 | 沈阳民航东北凯亚有限公司 | Fare prediction method and device |
| CN111028020A (en) * | 2019-12-17 | 2020-04-17 | 上海携程商务有限公司 | Yield management method, system, medium and electronic device for flight bag machine scene |
| CN112541629A (en) * | 2020-12-10 | 2021-03-23 | 携程旅游网络技术（上海）有限公司 | Prediction method, system, electronic device and storage medium for airport selection |
| CN115809903A (en) * | 2023-02-06 | 2023-03-17 | 北京四方启点科技有限公司 | Ticket buying method and device, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20240053715A1 (en) | Building energy storage system with planning tool | |
| López et al. | Parsimonious short-term load forecasting for optimal operation planning of electrical distribution systems | |
| CN109472399A (en) | Consider the air ticket purchase decision method and system of uncertainty in traffic | |
| Singh et al. | Short term load forecasting using artificial neural network | |
| Friedrich et al. | Short-term forecasting of the Abu Dhabi electricity load using multiple weather variables | |
| EP3447258A1 (en) | Central plant control system with equipment maintenance evaluation | |
| Robert et al. | Low cost, highly reliable rural electrification through a combination of grid extension and local renewable energy generation | |
| Feinberg et al. | Load forecasting | |
| Lin et al. | A privacy-preserving federated learning method for probabilistic community-level behind-the-meter solar generation disaggregation | |
| Khatib | Integrating project risk management and value engineering in tendering processes | |
| Larscheid et al. | Potential of new business models for grid integrated water electrolysis | |
| Beigaite et al. | Electricity price forecasting for nord pool data | |
| Ghadimi et al. | Short-term management of hydro-power systems based on uncertainty model in electricity markets | |
| JP6445909B2 (en) | Power demand forecasting system | |
| Janekova et al. | Assessment of economic efficiency and risk of the project using simulation. | |
| Ozoh et al. | A comparative analysis of techniques for forecasting electricity consumption | |
| BROŻYNA et al. | MULTI-SEASONALITY IN THE TBATS MODEL USING DEMAND FOR ELECTRIC ENERGY AS A CASE STUDY. | |
| Evangelopoulos et al. | Probabilistic spatial load forecasting for assessing the impact of electric load growth in power distribution networks | |
| KR20140069737A (en) | power bidding system and method for bidding thereof | |
| Arasteh et al. | Toward a smart distribution system expansion planning by considering demand response resources | |
| Dalal et al. | Day-ahead forecasting of losses in the distribution network | |
| Koivisto et al. | Statistical modeling of aggregated electricity consumption and distributed wind generation in distribution systems using AMR data | |
| Marinescu et al. | A dynamic forecasting method for small scale residential electrical demand | |
| Ding et al. | Individual nonparametric load estimation model for power distribution network planning | |
| Ghosh | Univariate time-series forecasting of monthly peak demand of electricity in northern India |
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: 20190315 |