CN107153888A - A kind of optimization Address Selection of Chain Store method based on extreme learning machine - Google Patents
A kind of optimization Address Selection of Chain Store method based on extreme learning machine Download PDFInfo
- Publication number
- CN107153888A CN107153888A CN201710283247.2A CN201710283247A CN107153888A CN 107153888 A CN107153888 A CN 107153888A CN 201710283247 A CN201710283247 A CN 201710283247A CN 107153888 A CN107153888 A CN 107153888A
- Authority
- CN
- China
- Prior art keywords
- feature
- learning machine
- extreme learning
- region
- data
- 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
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
- G06Q30/0204—Market segmentation
- G06Q30/0205—Location or geographical consideration
Abstract
The invention discloses a kind of optimization Address Selection of Chain Store method based on extreme learning machine, including:First according to city road network zoning, then using the social media in each region and sensing data construction feature, and limit of utilization automatic coding machine carries out Fusion Features.Finally according to the variance factor of data distribution between city, stack extreme learning machine and adaptive field extreme learning machine are flexibly used, training optimizes Address Selection of Chain Store method.Limit of utilization automatic coding machine of the present invention is merged to different views data, can effectively extend the data in other sources, while the present invention is based on adaptive field extreme learning machine technology, preferable effect can be also obtained to the small city possessed compared with small sample.
Description
Technical field
The invention belongs to data mining and city calculating field, and in particular to a kind of optimization based on extreme learning machine connects
Locksmith site selecting method.
Background technology
Retail shop, which optimizes addressing, can bring very strong economic benefit.The addressing of optimization is more usual than random addressing to be attracted more
Many customers.Such as one new cafe can be opened near the intersection of road, and usual intersection possesses easily traffic
With preferable passenger flow, but the factor such as traffic congestion be able to may also be negatively affected to this addressing.Calculated with city
Continue to develop, the optimization addressing for carrying out retail shop of retail shop using the mass data in city becomes possibility.Traditional optimization
Addressing generally builds model using a region feature itself such as flow of the people, purchasing power, traffic.But people not necessarily can
Some specific regions are rested on always.For example in big city, there are morning groups of people to be gone to work from far suburbs to downtown,
People is in certain flowing.The function (being for example commercial circle) that one region possesses in itself can also be produced to the quality of addressing
Influence.In addition, the competition factor that the same type of retail shop in region is brought is also required to be worthy of consideration.
One optimal addressing, which should be one, can attract the region of most users.In view of factor in summary,
It is that can set up out optimal site selection model generally to gather corresponding data in big city.However, some of city region
Or for some small cities, partial data is quite sparse such as social media, individually with these regions or city sheet
The model that the data of body are set up can not obtain good effect.In addition, the mode of traditional learning model needs to optimize magnanimity
Parameter, training effectiveness is low.In recent years, extreme learning machine obtains higher results of learning as one kind in mass data study
Learning framework, achieve prominent achievement in many fields.Stack extreme learning machine can be good at reply high dimensional feature and magnanimity
Data sample learning tasks, adaptive field extreme learning machine can using different distributions data and obtain preferable
Practise effect.Therefore, extreme learning machine is very feasible for carrying out chain store to optimize location problem using magnanimity Urban Data.
The content of the invention
In view of this, the invention provides a kind of optimization Address Selection of Chain Store method based on extreme learning machine.Compared to it
His method, the present invention realizes the optimization siting analysis in city sparse to data volume and abundant, is imitated with higher study
Rate, and cost is more cheap.
A kind of optimization Address Selection of Chain Store method based on extreme learning machine, comprises the following steps:
(1) using the road network in each city as border, city is divided into several adjacent regions;
(2) the social media data and physical sensors data composition data sample in each region are gathered, and utilize number
According to social media feature related to addressing in each region of sample architecture and physical sensors feature composition characteristic;
(3) social media data and physical sensors data are regarded as to the data of different views, and utilize own coding algorithm
Social media feature and physical sensors feature in the same period in each region is merged, fusion feature is obtained;
(4) selection needs to optimize the city of retail shop's addressing as target cities, chooses relative to target cities sample number
Source city is constructed according to the more cities of characteristic,
If the feature relative entropy of target cities and source city is less than threshold value, by melting for each region in source city
Feature is closed as the input of stack extreme learning machine, stack extreme learning machine is trained, obtains site selection model;
Otherwise, learn the fusion feature in each region in target cities and source city as the adaptive field limit
The input of machine, trains adaptive field extreme learning machine, obtains site selection model;
(5) each region in target cities is tested using site selection model, obtains optimal addressing region.
In step (2), described social media data refer to obtaining from microblogging, popular comment and other social medias
Social media text, those data can clearly reflect emotion of the people to region, can be used as relatively effective addressing
Data.Described physical sensors data refer to obtaining number from traffic, bus, property price, point of interest and commercial circle
According to.
In step (2), to analyze term vector, word frequency rate and the user of the acquisition of social media text to retail shop in region
Environment, the scoring of service are used as social media feature.
In step (2), to calculate daily (6 points~10 points, 11 points~15 points, 16 points of different time sections in obtained region
~20 points) traffic average is used as traffic characteristic;Public transport is used as to calculate public transport vehicle shift number and vehicle flowrate in obtained region
Feature;Room rate feature is used as to calculate room rate average in obtained region;The retail shop to be opened is same in the region obtained with calculating
Class retail shop and total retail shop's ratio are used as competition feature;Traffic characteristic, public transport feature, room rate feature and competition feature composition physics
Sensor characteristics.
In step (3), using social media feature and physical sensors feature as different views, using limit own coding
Machine merges social media feature and physical sensors feature, and detailed process is:
First, the first layer weight of the self-editing ink recorder of the random initializtion limit and bigoted;
Then, the weight of follow-up hidden layer is tried to achieve and bigoted by optimizing square difference function exported between input;
Finally, the random weight of first layer, the bigoted weight, bigoted by social media feature and thing tried to achieve with optimization are utilized
Reason sensor characteristics do nonlinear transformation, export fusion feature.
In step (4), the feature relative entropy of target cities and source city is smaller, represent sample data distributional difference compared with
It is small, typically it is probably spatially close between city, now, regard the fusion feature in each region in source city as stack pole
Limit learning machine input, conversely, using the data of target cities and source city as different field data, now, by target
The fusion feature in each region in city and source city as adaptive field extreme learning machine input.
In step (4), relative entropy is also known as KL divergences, measurement be two probability distribution in similar events space difference
Situation.Its physical significance is:In similar events space, probability distribution P (x) event space, if being compiled with probability distribution Q (x)
During code, average each elementary event x code lengths add how many bit.KL distances are represented with D (P | | Q), calculation formula is such as
Under:
When two probability distribution are identical, i.e. P (x)=Q (x), its relative entropy is that 0, X is the set of elementary event, this hair
In bright, the span with respect to entropy threshold is 0~0.2.
In step (4), for stack extreme learning machine, the mode of many hidden layers is taken to obtain defeated using multiple nonlinear transformation
Go out value, the weight of usual first layer and bigoted to be random.For adaptive field extreme learning machine, the side of single hidden layer is taken
Formula obtains output valve using 1 nonlinear transformation;The loss function of stack extreme learning machine and adaptive field extreme learning machine is all
It is square difference function between each layer of output and input.
In step (5), each region in target cities is tested using site selection model, optimal addressed area is obtained
The detailed process in domain is:
First, the following attraction volume of the flow of passengers probable value for obtaining target cities is calculated using site selection model;
Then, the volume of the flow of passengers probable value to all regions in target cities carries out drop minor sort, selects volume of the flow of passengers probable value
N region is as optimal addressing region before coming, and n represents natural number.
Traditional optimizes siting analysis to retail shop, has great lack in face of less sample or less characteristic
Fall into;And the present invention is based on extreme learning machine, it is possible to use adaptive field extreme learning machine is modeled.Specific advantage embodies
It is as follows:
(1) magnanimity Urban Data is faced, this method learning efficiency is higher, and pace of learning is faster than regular machinery learning algorithm.
(2) this method is applied to not only possess the city of abundant sample data, can also for the sparse city of sample
Preferable effect is obtained using domain-adaptive and metropolitan data.
Brief description of the drawings
Fig. 1 is the optimization Address Selection of Chain Store method flow schematic diagram of the invention based on extreme learning machine;
Fig. 2 is that the self-editing ink recorder of the different pieces of information view limit of the present invention carries out Fusion Features schematic diagram.
Embodiment
In order to more specifically describe the present invention, below in conjunction with the accompanying drawings and embodiment is to technical scheme
It is described in detail.
As shown in figure 1, the optimization Address Selection of Chain Store method based on extreme learning machine, including:
S01, using the road network in each city as border, several adjacent regions are divided into by city.
S02, gathers the social media data and physical sensors data composition data sample in each region, and utilize number
According to social media feature related to addressing in each region of sample architecture and physical sensors feature composition characteristic.
Social media data are the social media texts obtained from microblogging, popular comment, are obtained with analyzing social media text
Term vector, word frequency rate and the user obtained is used as social media feature to the environment of retail shop, the scoring serviced in region.
Physical sensors data are to obtain data from traffic, bus, property price, point of interest and commercial circle, in terms of
Daily different time sections (6. -10 points, 11. -15 points, 16. -20 points) traffic average is used as friendship in obtained region
Logical feature;Public transport feature is used as to calculate public transport vehicle shift number and vehicle flowrate in obtained region;To calculate room in obtained region
Valency average is used as room rate feature;Competition is used as to calculate the similar retail shop for the retail shop to be opened and total retail shop's ratio in obtained region
Feature;Traffic characteristic, public transport feature, room rate feature and competition feature composition physical sensors feature.
S03, it is each using the self-editing ink recorder of the limit using social media feature and physical sensors feature as different views
Social media feature and physical sensors feature in region in the same period are merged, and detailed process is:
First, the first layer weight of the self-editing ink recorder of the random initializtion limit and bigoted;
Then, the weight of follow-up hidden layer is tried to achieve and bigoted by optimizing square difference function exported between input;
Finally, the random weight of first layer, the bigoted weight, bigoted by social media feature and thing tried to achieve with optimization are utilized
Reason sensor characteristics do nonlinear transformation, export fusion feature.
S04, selection needs to optimize the city of retail shop's addressing as target cities, chooses relative to target cities sample number
Source city is constructed according to the more cities of characteristic.
S05, judges whether the feature relative entropy of target cities and source city is less than threshold value, if so, S06 is performed, if it is not,
Perform S07.
S06, using the fusion feature in each region in source city as the input of stack extreme learning machine, trains the stack limit
Learning machine, obtains site selection model.
S07 regard the fusion feature in each region in target cities and source city as adaptive field extreme learning machine
Input, train adaptive field extreme learning machine, obtain site selection model.
S08, is tested each region in target cities using site selection model, obtains optimal addressing region.
Detailed process is:
First, the following attraction volume of the flow of passengers probable value for obtaining target cities is calculated using site selection model;
Then, the volume of the flow of passengers probable value to all regions in target cities carries out drop minor sort, selects volume of the flow of passengers probable value
N region is as optimal addressing region before coming, and n represents natural number.
Embodiment 1
Selection Beijing, Shanghai, Guangzhou, Wuhan, Shenzhen are source city, and Hangzhou is the target cities of addressing.According to Hangzhou
Road network situation, be divided into the region of 4315 to Hangzhou, and each region is that favored area is treated in addressing.
Analysis can be obtained, the data characteristics Normal Distribution in this five source city features and Hangzhou, i.e., with Hangzhou feature
Relative entropy is less than threshold value 0.1, and the fusion feature in each region in Hangzhou, Beijing, Shanghai, Guangzhou, Wuhan and Shenzhen is made
For the input of adaptive field extreme learning machine, adaptive field extreme learning machine is trained, site selection model is obtained;Next utilize
Site selection model obtains 5 regions as optimal addressing region.
Technical scheme and beneficial effect are described in detail above-described embodiment, Ying Li
Solution is to the foregoing is only presently most preferred embodiment of the invention, is not intended to limit the invention, all principle models in the present invention
Interior done any modification, supplement and equivalent substitution etc. are enclosed, be should be included in the scope of the protection.
Claims (8)
1. a kind of optimization Address Selection of Chain Store method based on extreme learning machine, comprises the following steps:
(1) using the road network in each city as border, city is divided into several adjacent regions;
(2) the social media data and physical sensors data composition data sample in each region are gathered, and utilize data sample
The social media feature related to addressing and physical sensors feature composition characteristic in each region of this construction;
(3) social media data and physical sensors data are regarded as to the data of different views, and using own coding algorithm to every
Social media feature and physical sensors feature in individual region in the same period are merged, and obtain fusion feature;
(4) selection needs to optimize the city of retail shop addressing as target cities, choose relative to target cities sample data and
The more city construction source cities of characteristic,
It is if the feature relative entropy of target cities and source city is less than threshold value, the fusion in each region in source city is special
The input as stack extreme learning machine is levied, stack extreme learning machine is trained, obtains site selection model;
Otherwise, it regard the fusion feature in each region in target cities and source city as adaptive field extreme learning machine
Input, trains adaptive field extreme learning machine, obtains site selection model.
(5) each region in target cities is tested using site selection model, obtains optimal addressed areas.
2. the optimization Address Selection of Chain Store method as claimed in claim 1 based on extreme learning machine, it is characterised in that described
Social media data refer to the social media text obtained from microblogging, popular comment.
3. the optimization Address Selection of Chain Store method as claimed in claim 1 based on extreme learning machine, it is characterised in that described
Physical sensors data refer to the data obtained from traffic, bus, property price, point of interest and commercial circle.
4. the optimization Address Selection of Chain Store method as claimed in claim 2 based on extreme learning machine, it is characterised in that to analyze
Term vector, word frequency rate and the user that social media text is obtained are used as social activity to the environment of retail shop, the scoring serviced in region
Media characteristic.
5. the optimization Address Selection of Chain Store method as claimed in claim 3 based on extreme learning machine, it is characterised in that to calculate
Daily different time sections traffic average is as traffic characteristic in obtained region;To calculate public transport vehicle shift in obtained region
Number and vehicle flowrate are used as public transport feature;Room rate feature is used as to calculate room rate average in obtained region;To calculate obtained area
The similar retail shop for the retail shop to be opened in domain and total retail shop's ratio are used as competition feature;Traffic characteristic, public transport feature, room rate feature with
And competition feature composition physical sensors feature.
6. the optimization Address Selection of Chain Store method as claimed in claim 5 based on extreme learning machine, it is characterised in that step
(3) in, using social media feature and physical sensors feature as different views, social matchmaker is merged using the self-editing ink recorder of the limit
Body characteristicses and physical sensors feature, detailed process is:
First, the first layer weight of the self-editing ink recorder of the random initializtion limit and bigoted;
Then, the weight of follow-up hidden layer is tried to achieve and bigoted by optimizing square difference function exported between input;
Finally, using the random weight of first layer, the bigoted weight tried to achieve with optimization, bigoted social media feature and physics are passed
Sensor feature does nonlinear transformation, exports fusion feature.
7. the optimization Address Selection of Chain Store method as claimed in claim 1 based on extreme learning machine, it is characterised in that step
(4) in, for stack extreme learning machine, the mode of many hidden layers is taken to obtain output valve, first layer using multiple nonlinear transformation
Weight and bigoted to be random;For adaptive field extreme learning machine, take the mode of single hidden layer non-linear using 1 time
Conversion obtains output valve;The loss function of stack extreme learning machine and adaptive field extreme learning machine be all each layer output with it is defeated
Square difference function between entering.
8. the optimization Address Selection of Chain Store method as claimed in claim 1 based on extreme learning machine, it is characterised in that step
(5) in, each region in target cities is tested using site selection model, the detailed process of optimal addressed areas is obtained
For:
First, the following attraction volume of the flow of passengers probable value for obtaining target cities is calculated using site selection model;
Then, the volume of the flow of passengers probable value to all regions in target cities carries out drop minor sort, and selection volume of the flow of passengers probable value comes
Preceding n region represents natural number as optimal addressing region, n.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710283247.2A CN107153888A (en) | 2017-04-26 | 2017-04-26 | A kind of optimization Address Selection of Chain Store method based on extreme learning machine |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710283247.2A CN107153888A (en) | 2017-04-26 | 2017-04-26 | A kind of optimization Address Selection of Chain Store method based on extreme learning machine |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107153888A true CN107153888A (en) | 2017-09-12 |
Family
ID=59793625
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710283247.2A Pending CN107153888A (en) | 2017-04-26 | 2017-04-26 | A kind of optimization Address Selection of Chain Store method based on extreme learning machine |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107153888A (en) |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108446800A (en) * | 2018-03-12 | 2018-08-24 | 西北工业大学 | A kind of site selecting method and system of trans-city chain |
CN109191181A (en) * | 2018-08-08 | 2019-01-11 | 北京工商大学 | A kind of digital signage commercial audience listener clustering method based on neural network and Huff model |
CN109658172A (en) * | 2017-12-13 | 2019-04-19 | 美味不用等(上海)信息科技股份有限公司 | A kind of commercial circle recommended method calculates unit and storage medium |
CN110111150A (en) * | 2019-05-08 | 2019-08-09 | 拉扎斯网络科技(上海)有限公司 | Information processing method, device, storage medium and electronic equipment |
CN110414751A (en) * | 2018-04-26 | 2019-11-05 | 观相科技(上海)有限公司 | A kind of hotel industry addressing evaluation system and evaluation method based on geographical location |
CN110889029A (en) * | 2018-08-17 | 2020-03-17 | 北京京东金融科技控股有限公司 | City target recommendation method and device |
CN111708859A (en) * | 2020-06-18 | 2020-09-25 | 中科宇图科技股份有限公司 | Business turn boundary identification method based on spatial data |
CN114500720A (en) * | 2022-01-10 | 2022-05-13 | 成都易达数安科技有限公司 | Media-integration parallel APP city switching method based on dynamic menu and column technology |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103729648A (en) * | 2014-01-07 | 2014-04-16 | 中国科学院计算技术研究所 | Domain adaptive mode identifying method and system |
CN105203869A (en) * | 2015-09-06 | 2015-12-30 | 国网山东省电力公司烟台供电公司 | Microgrid island detection method based on extreme learning machine |
CN105512679A (en) * | 2015-12-02 | 2016-04-20 | 天津大学 | Zero sample classification method based on extreme learning machine |
WO2017007742A1 (en) * | 2015-07-06 | 2017-01-12 | Microsoft Technology Licensing, Llc | Transfer learning techniques for disparate label sets |
CN106569954A (en) * | 2016-11-08 | 2017-04-19 | 南京航空航天大学 | Method based on KL divergence for predicting multi-source software defects |
-
2017
- 2017-04-26 CN CN201710283247.2A patent/CN107153888A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103729648A (en) * | 2014-01-07 | 2014-04-16 | 中国科学院计算技术研究所 | Domain adaptive mode identifying method and system |
WO2017007742A1 (en) * | 2015-07-06 | 2017-01-12 | Microsoft Technology Licensing, Llc | Transfer learning techniques for disparate label sets |
CN105203869A (en) * | 2015-09-06 | 2015-12-30 | 国网山东省电力公司烟台供电公司 | Microgrid island detection method based on extreme learning machine |
CN105512679A (en) * | 2015-12-02 | 2016-04-20 | 天津大学 | Zero sample classification method based on extreme learning machine |
CN106569954A (en) * | 2016-11-08 | 2017-04-19 | 南京航空航天大学 | Method based on KL divergence for predicting multi-source software defects |
Non-Patent Citations (2)
Title |
---|
NINGYU ZHANG: "《ELM Meets Urban Big Data Analysis:Case Studies》", 《COMPUTATIONAL INTELLIGENCE AND NEUROSCIENCE》 * |
NINGYU ZHANG: "《Transfer Learning for Urban Computing:A case study for Optimal Retail Store Placement》", 《URBCOMP"15》 * |
Cited By (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109658172A (en) * | 2017-12-13 | 2019-04-19 | 美味不用等(上海)信息科技股份有限公司 | A kind of commercial circle recommended method calculates unit and storage medium |
CN108446800A (en) * | 2018-03-12 | 2018-08-24 | 西北工业大学 | A kind of site selecting method and system of trans-city chain |
CN110414751A (en) * | 2018-04-26 | 2019-11-05 | 观相科技(上海)有限公司 | A kind of hotel industry addressing evaluation system and evaluation method based on geographical location |
CN109191181A (en) * | 2018-08-08 | 2019-01-11 | 北京工商大学 | A kind of digital signage commercial audience listener clustering method based on neural network and Huff model |
CN109191181B (en) * | 2018-08-08 | 2021-08-27 | 北京工商大学 | Digital signage advertisement audience and crowd classification method based on neural network and Huff model |
CN110889029A (en) * | 2018-08-17 | 2020-03-17 | 北京京东金融科技控股有限公司 | City target recommendation method and device |
CN110889029B (en) * | 2018-08-17 | 2024-04-05 | 京东科技控股股份有限公司 | Urban target recommendation method and device |
CN110111150A (en) * | 2019-05-08 | 2019-08-09 | 拉扎斯网络科技(上海)有限公司 | Information processing method, device, storage medium and electronic equipment |
CN111708859A (en) * | 2020-06-18 | 2020-09-25 | 中科宇图科技股份有限公司 | Business turn boundary identification method based on spatial data |
CN111708859B (en) * | 2020-06-18 | 2023-04-25 | 中科宇图科技股份有限公司 | Business circle boundary identification method based on spatial data |
CN114500720A (en) * | 2022-01-10 | 2022-05-13 | 成都易达数安科技有限公司 | Media-integration parallel APP city switching method based on dynamic menu and column technology |
CN114500720B (en) * | 2022-01-10 | 2024-03-26 | 成都易达数安科技有限公司 | Melt media parallel APP city switching method based on dynamic menu and column technology |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107153888A (en) | A kind of optimization Address Selection of Chain Store method based on extreme learning machine | |
CN106952105A (en) | A kind of retail shop based on transfer learning optimizes site selecting method | |
Wang et al. | Driving factors of urban shrinkage: Examining the role of local industrial diversity | |
Petrov et al. | Urban land use scenarios for a tourist region in Europe: Applying the MOLAND model to Algarve, Portugal | |
Cheng et al. | Developing an evaluation index system for low-carbon tourist attractions in China–A case study examining the Xixi wetland | |
CN108022012A (en) | Vehicle location Forecasting Methodology based on deep learning | |
CN108133302A (en) | A kind of public bicycles potential demand Forecasting Methodology based on big data | |
Xie et al. | Spatial logistic regression and GIS to model rural-urban land conversion | |
Hersperger et al. | Driving forces of landscape change in the urbanizing Limmat valley, Switzerland | |
Scott et al. | Network analysis methods for modeling tourism inter-organizational systems | |
CN110414751A (en) | A kind of hotel industry addressing evaluation system and evaluation method based on geographical location | |
Abdel Azim Ahmed | A triple bottom line analysis of the impacts of the Hail International Rally in Saudi Arabia | |
Jones et al. | Mountainous protected areas & nature-based tourism in Asia | |
Ogorelc | Residents' perceptions of tourism impacts and sustainable tourism development | |
CN110119838A (en) | A kind of shared bicycle demand forecast system, method and device | |
Battarra et al. | Smart mobility: An evaluation method to audit Italian cities | |
Supriyadi et al. | The economic impact of international tourism to overcome the unemployment and the poverty in Indonesia | |
Wong et al. | An analysis of dual‐track urbanisation in the Pearl River Delta since 1980 | |
Moore | Getting serious about sustainability: Exploring the potential for One-Planet Living in Vancouver | |
Bouachera et al. | Fuel demand and car ownership modelling in India | |
Abu et al. | SARIMA and Exponential Smoothing model for forecasting ecotourism demand: A case study in National Park Kuala Tahan, Pahang | |
Khan et al. | Analysis of Ecological Footprint of Rural-Urban Households in Islamabad | |
Sharma et al. | Sustainable tourism using decision support system based on system dynamics: A case study from Amsterdam | |
Chan et al. | Galvanizing the groundswell of climate actions in the developing world | |
Straatman et al. | A consumption-based, regional input-output analysis of greenhouse gas emissions and the carbon regional index |
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: 20170912 |