CN115860812A - Brand store entrance rate prediction method and system based on data lake - Google Patents
Brand store entrance rate prediction method and system based on data lake Download PDFInfo
- Publication number
- CN115860812A CN115860812A CN202211548984.8A CN202211548984A CN115860812A CN 115860812 A CN115860812 A CN 115860812A CN 202211548984 A CN202211548984 A CN 202211548984A CN 115860812 A CN115860812 A CN 115860812A
- Authority
- CN
- China
- Prior art keywords
- store
- brand
- data
- model
- information
- 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
- 238000000034 method Methods 0.000 title claims abstract description 27
- 238000012549 training Methods 0.000 claims abstract description 29
- 238000009826 distribution Methods 0.000 claims abstract description 14
- 230000007774 longterm Effects 0.000 claims abstract description 7
- 238000012545 processing Methods 0.000 claims description 17
- 230000006870 function Effects 0.000 claims description 11
- 238000012360 testing method Methods 0.000 claims description 10
- 230000002093 peripheral effect Effects 0.000 claims description 9
- 238000007476 Maximum Likelihood Methods 0.000 claims description 8
- 238000010276 construction Methods 0.000 claims description 7
- 230000007613 environmental effect Effects 0.000 claims description 7
- 230000002159 abnormal effect Effects 0.000 claims description 6
- 238000004140 cleaning Methods 0.000 claims description 6
- 238000001514 detection method Methods 0.000 claims description 6
- 238000005457 optimization Methods 0.000 claims description 6
- 238000013135 deep learning Methods 0.000 claims description 5
- 238000010801 machine learning Methods 0.000 claims description 5
- 238000012795 verification Methods 0.000 claims description 5
- 238000006243 chemical reaction Methods 0.000 claims description 4
- 230000000694 effects Effects 0.000 claims description 4
- 238000005516 engineering process Methods 0.000 claims description 4
- 230000002776 aggregation Effects 0.000 claims description 3
- 238000004220 aggregation Methods 0.000 claims description 3
- 230000002860 competitive effect Effects 0.000 claims description 3
- 230000008030 elimination Effects 0.000 claims description 3
- 238000003379 elimination reaction Methods 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 abstract 1
- 238000013461 design Methods 0.000 description 2
- 230000010365 information processing Effects 0.000 description 2
- 230000004931 aggregating effect Effects 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000008034 disappearance Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a data lake-based store entering rate prediction method and system for brand stores, which specifically comprise the following steps: the method comprises the steps of collecting off-line data required by model training, storing the off-line data in a data lake in a centralized mode, conducting data preprocessing on the collected data, setting a prediction target dimension, constructing samples and characteristics according to the dimension, finding out an optimal model and parameters through model automatic modeling, inputting all samples into the optimal model to obtain a final on-line model, predicting the store-entering rate of the samples to be predicted on line, and estimating the long-term and short-term store-entering rates of brand stores at the planned-operation position, so that a foundation is laid for further estimating the store-opening benefits. The present invention solves this problem by modeling the store entry rate using binomial distributions. The store-entering rate model of the system provided by the invention is constructed by collecting different store data of different brands into a data lake, and each brand is called and used through an online API (application program interface).
Description
Technical Field
The invention relates to the technical field of artificial intelligence, in particular to a data lake-based store entering rate prediction method and system for brand stores.
Background
Along with social development, the e-commerce becomes an infrastructure, the threshold of online opening is low and flexible, but with the disappearance of online flow red profit, the online growth has a bottleneck, and online marketing cost is increasing day by day, consumers also pay more and more attention to offline experience consumption, so the offline flash shop and slow flash shop form with flexible opening as the characteristics is created, different brands return more and more to offline opening, experience consumption is brought to users, and through the flexible opening form, brands can rapidly reach their target consumption groups by opening shops in different places.
During the process of opening a store, the sales of stores at a certain planned-to-be-opened position are estimated in advance according to the formula of passenger flow rate, store-entering rate, conversion rate and customer price, and then the ROI is calculated by combining the cost to determine whether to open the store at the position. Traditionally, different brands estimate the store-in rate of a store to be operated according to the historical average of store-passing and store-entering passenger flow data of the existing stores, or estimate a store-in rate according to past experience. In fact, the store-in rate of different brands at different positions is different, the traditional method does not consider the difference, and the estimation of the store-in rate is not accurate.
In addition, the existing store-entering rate estimation model generally directly takes the store-entering rate as a prediction target to construct a model, and does not consider the property that the store-entering rate is in the interval of [ 0,1 ], which possibly causes the prediction result of the model to violate the property.
Therefore, aiming at the problem of estimating the store-in rate of different brands at different positions, the invention provides a store-in rate prediction method and a store-in rate prediction system based on binomial distribution modeling by integrating multiple influence factors related to brands and store positions, and helps the different brands to effectively estimate the store-in rate when the stores are planned to be opened at different positions and support store-opening decisions.
Disclosure of Invention
Aiming at the problem that the store entering rate cannot be accurately predicted in the prior art, the invention provides a store entering rate prediction method and system for brand stores based on a data lake.
The invention provides a data lake-based store entering rate prediction method for brand stores, which comprises the following steps: comprises that
S1, collecting historical data: the system comprises necessary modeling data such as Store information Store (poi), location information Loc _ info (place), place information (place), peripheral information Sur _ info (place), trade Brand information Brand _ info (Brand), environment information Context, passenger Flow data Flow (poi) and the like;
and S2, data processing: carrying out data cleaning, abnormal value detection and elimination, and missing value filling; s3, sample construction and model training: constructing characteristics for the processed data, and encoding the category characteristics; modeling the store-entering rate based on binomial distribution, and constructing a maximum likelihood function as an optimization target to solve model parameters; taking samples in a certain proportion from all samples, carrying out model training and parameter adjustment of machine learning or deep learning to obtain an optimal model, and retraining the model on-line deployment by using all samples;
s4, online prediction: and calling the trained model to obtain an online prediction result according to the collected characteristic information required by the model of the planned store.
Preferably, in step S1:
store information Store (poi) is collected: entering store information to be collected, including specific information such as brands, specific locations of stores, store types, position types, places (specific market names, ids or building names, ids), floors, house numbers, areas, time of opening, rent and the like, from a store information management system;
collection position information Loc _ info (place): the method comprises the basic position information of province and city areas, longitude and latitude, business circles, city grades, city types and the like;
collecting site information Place _ info (Place): the method comprises the following steps of including site addresses, site types, site ids, site areas, building time, site floors, site average rent, site brand preference, passenger flow level, consumption level and other site label drawings, site business state distribution, site passenger group crowd drawings, site historical transaction information, site floor business state distribution and the like, wherein the site types relate to shopping malls, office buildings, scenic spots, street pavements, communities and the like, and the sites refer to names and numbers of specific shopping centers, office buildings, communities, buildings and the like; collecting peripheral information Sur _ info (place) including transportation facilities and public facilities (gas stations, charging stations, transportation facilities, and the like) around a site;
the collected environment information Context comprises weather conditions, temperature, season, holidays, weeks, promotion activities and the like;
acquiring Brand information Brand _ info (Brand) of a merchant, wherein the Brand information Brand _ info (Brand) comprises industries, enterprises, numbers of chain stores, brand positioning, target customer groups, product average price, competitive product brands and the like;
the collected passenger Flow data Flow (poi) comprises the number of people entering the store, the number of people passing the store, the collection time, the passenger group attributes (such as sex ratio, age group ratio and passenger Flow direction ratio);
preferably, the passenger flow data in step S1 includes:
the method comprises the steps that through passenger flow detection equipment comprising a fixed camera and mobile video shooting equipment, the number of people entering a store, the number of people passing the store and corresponding data acquisition time of the number of people are obtained by applying a human body recognition and target tracking technology, and data are sent to a back-end database; and collecting the passenger flow data of different shops of different brands together to form a lake of the passenger flow data of the shops.
Preferably, in step S2: by adopting the promotion environment information processing historical data in the step S1, the promotion days can be eliminated, or the increase rate of the promotion days relative to the daily store-in rate can be calculated, so that the store-in rate of the promotion days is reduced to the daily.
Preferably, in step S3: because the future environmental information is unknown, a prediction target and a sample with specified dimensions can be respectively set according to a business target, a model is built according to different dimensions, the dimensions are divided according to multiple dimensions such as months, weekday weekends and the like, an average store-in rate prediction target of stores (poi) with dimensions (month and date types) is formed, and long-term and short-term prediction tasks are realized.
Preferably, the data after processing is characterized by specifically: after data are aggregated through specified dimensions, characteristics required by the model are generated through characteristic conversion, characteristic coding and characteristic combination. The environmental information characteristics are not preserved after aggregation.
Preferably, based on the store-entering rate modeling of the binomial distribution, a maximum likelihood function is constructed to be solved as an optimization objective function, as follows:
wherein, T i Number of total store-passing persons representing ith time period of brand store, N i Indicating the total number of people entering the store, X i Characteristic factors influencing the store-in rate of brand stores, store-in rate
w is a parameter that needs to be estimated by the sample. Pi (X) i ) Other forms are possible depending on the particular problem. The time period division may be a day or a week or a month, assuming a total of N segments, i.e. N samples. The problem of parameter w estimation is to find the optimal w to maximize L (w).
Preferably, in step S3: and finally, selecting a model with the optimal performance of the test set to retrain all samples to obtain a model to be on-line.
A data lake based store entry rate prediction system for a brand store, comprising:
the data acquisition module is used for acquiring basic data required by model training, and the basic data comprises Store information Store (poi), location information Loc _ info (place), site information (place), peripheral information Sur _ info (place), merchant Brand information Brand _ info (Brand), environment information Context and passenger Flow data Flow (poi);
the data processing module is used for processing and cleaning the collected data, processing abnormal values and the like and removing noise;
and the sample construction module is used for generating samples and characteristics required by the model, and the module also comprises a prediction target dimension determination unit which constructs the target and the sample according to the set dimension.
The model training module is used for inputting the training samples into the model to obtain a store-entering rate prediction model, and further comprises a sample dividing unit which is used for dividing a training set, a verification set and a test set, a model training parameter searching unit, an optimal model obtained by various algorithms, a model training unit to be on-line and a model to be on-line, and all samples are input into the optimal model to obtain a final model to be on-line.
And the online prediction module is used for evaluating the short-term and long-term future store-entering rate of the planned-operation position brand, inputting the characteristics required by the model and obtaining a store-entering rate prediction result.
The invention has the beneficial effects that:
compared with the prior art, the method provided by the invention fully utilizes big data of various dimensions, adopts machine learning and deep learning automatic modeling technologies to scientifically predict the store entrance rate of different brands at different positions, and changes the current simple measurement and calculation according to the conventional stores. When more and more stores and brands are accumulated in the system, the model prediction result is more and more accurate, and the matching degree estimation capability of the brands and the positions can be improved more and more.
Drawings
FIG. 1 is a flow chart showing a specific method employed by an embodiment of the present invention;
fig. 2 is an exemplary diagram of a system in which embodiments of the present invention may be employed.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings, but the present invention is not limited thereto.
In an embodiment of the present invention, referring to fig. 1:
s1, collecting historical data: the system comprises necessary modeling data such as Store information Store (poi), location information Loc _ info (place), place information (place), peripheral information Sur _ info (place), merchant Brand information Brand _ info (Brand), environment information Context, passenger Flow data Flow (poi) and the like;
and S2, data processing: carrying out data cleaning, abnormal value detection and elimination and missing value filling;
s3, sample construction and model training: constructing characteristics of the processed data, and encoding the category characteristics; modeling the store-entering rate based on binomial distribution, and constructing a maximum likelihood function as an optimization target to solve model parameters; taking samples in a certain proportion from all samples, carrying out model training and parameter adjustment of machine learning or deep learning to obtain an optimal model, and retraining the model on-line deployment by using all samples;
and S4, online prediction: and calling the trained model to obtain an online prediction result according to the collected characteristic information required by the model of the planned store.
In step S1:
store information Store (poi): entering store information to be collected, including specific information such as brands, specific locations of stores, store types, position types, places (specific market names, ids or building names, ids), floors, house numbers, areas, time of opening, rent and the like, from a store information management system;
collection position information Loc _ info (place): the method comprises the basic position information of province and city areas, longitude and latitude, business circles, city grades, city types and the like;
collecting site information Place _ info (Place): the method comprises the following steps of (1) site address, site type, site id, site area, building time, site floor, site average rent, site brand preference, passenger flow level, consumption level and other site label drawings, site business state distribution, site customer group crowd drawings, site historical trading information, site floor business state distribution and the like, wherein the site type relates to shopping malls, office buildings, scenic spots, street pavements, communities and the like, and the site refers to names and numbers of specific shopping centers, office buildings, communities, buildings and the like; collecting peripheral information Sur _ info (place) including transportation facilities and public facilities (gas stations, charging stations, transportation facilities and the like) around a site;
the collected environment information Context comprises weather conditions, temperature, season, holidays, weeks, promotion activities and the like;
acquiring Brand information Brand _ info (Brand) of a merchant, wherein the Brand information Brand _ info (Brand) comprises industries, enterprises, numbers of chain stores, brand positioning, target customer groups, product average price, competitive Brand brands and the like;
the collected passenger Flow data Flow (poi) comprises the number of people entering the store, the number of people passing the store, the collection time, the passenger group attributes (such as sex ratio, age group ratio and passenger Flow direction ratio);
the passenger flow data in step S1 includes:
through passenger flow detection equipment comprising a fixed camera and mobile video shooting equipment, acquiring the number of people entering a store, the number of people passing the store and corresponding data acquisition time thereof by using a human body recognition and target tracking technology, and sending data to a back-end database; collecting the passenger flow data of different shops of different brands together to form a shop passenger flow data lake;
the step S2 comprises the following steps: by adopting the historical data of sales promotion and other environmental information processing in the step S1, the processing mode can eliminate sales promotion days or calculate the increase rate of sales promotion days relative to the daily store-in rate to reduce the store-in rate of sales promotion days to the daily,
the step S3 comprises the following steps: because the future environmental information is unknown, a prediction target and a sample of a designated dimension can be set according to a business target, a model is built according to different dimensions, the dimensions are divided according to a plurality of dimensions such as months, weekday weekends and the like, an average store-entering rate prediction target of stores (poi) of dimensions such as (month and date types) is formed, and long-term and short-term prediction tasks are realized. The dimensionality is not limited to two dimensionalities of a month and the end of a working day and week, and can be set and modified according to a prediction scene of a specific store-entering rate, so that the store-entering rate of stores under various conditions is comprehensively evaluated, and the comprehensive evaluation of the benefits of the stores is facilitated;
and (3) constructing the features, aggregating the data according to the claim 2 through the specified dimensions, and generating the features required by the model through feature conversion, feature coding and feature combination. The environmental information characteristics are not preserved after aggregation.
The store-entering rate modeling claim based on the binomial distribution establishes a maximum likelihood function as an optimization objective function to solve, and the maximum likelihood function is as follows:
wherein, T i Number of total store-passing persons representing ith time period of brand store, N i Indicates the total number of store entries, X i Characteristic factor for showing influence on store-in rate of brand stores, store-in rate
w is a parameter that needs to be estimated from the samples. Pi (x) i ) Other forms are possible depending on the particular problem. The time period division may be a day or a week or a month, assuming a total of N segments, i.e. N samples. The problem of parameter w estimation is to find the optimal w to maximize L (w).
The step S3 comprises the following steps: and finally, selecting a model with the optimal performance of the test set to retrain all samples to obtain a model to be on-line. For example, training parameter adjusting models A1, A2 and A3 \8230respectively, entering a model A1 with the optimal test effect in a test set, and then retraining the model A1 and all samples to obtain a model A which can be finally on-line.
A data lake based store entry rate prediction system for a brand store, comprising:
the data acquisition module is used for acquiring basic data required by model training, and the basic data comprises Store information Store (poi), location information Loc _ info (place), site information (place), peripheral information Sur _ info (place), merchant Brand information Brand _ info (Brand), environment information Context and passenger Flow data Flow (poi);
the data processing module is used for processing and cleaning the collected data, processing abnormal values and the like and removing noise;
and the sample construction module is used for generating samples and characteristics required by the model, and the module also comprises a prediction target dimension determination unit which constructs the target and the sample according to the set dimension.
The model training module is used for inputting the training samples into the model to obtain a store-entering rate prediction model, and further comprises a sample dividing unit which is used for dividing a training set, a verification set and a test set, a model training parameter searching unit, an optimal model obtained by various algorithms, a model training unit to be on-line and a model to be on-line, and all samples are input into the optimal model to obtain a final model to be on-line.
And the online prediction module is used for evaluating the short-term and long-term future store-entering rate of the planned-operation position brand, inputting the characteristics required by the model and obtaining a store-entering rate prediction result.
The present invention is not limited to the above-described embodiments, and those skilled in the art can implement the present invention in other various embodiments based on the disclosure of the present invention. Therefore, the design of the invention falls into the scope of protection when the design structure and thought of the invention are adopted and some simple changes or modifications are made.
Claims (9)
1. A store-entering rate prediction method of brand stores based on data lakes is characterized by comprising the following steps:
s1, collecting historical data: the system comprises Store information Store (poi), location information Loc _ info (place), place information (place), peripheral information Sur _ info (place), merchant Brand information Brand _ info (Brand), environment information Context and passenger Flow data Flow (poi) modeling essential data;
and S2, data processing: carrying out data cleaning, abnormal value detection and elimination and missing value filling;
s3, sample construction and model training: constructing characteristics of the processed data, and encoding the category characteristics; modeling the store-entering rate based on binomial distribution, and constructing a maximum likelihood function as an optimization target to solve model parameters; taking samples in a certain proportion from all samples, carrying out model training and parameter adjustment of machine learning or deep learning to obtain an optimal model, and retraining the model on-line deployment by using all samples;
and S4, online prediction: and performing online reasoning to obtain shop passing passenger flow information by stepping on points according to the collected characteristic information of the positions of the planned opening shops required by the model, and calling the trained model to obtain an online prediction result.
2. The method for predicting the store-in rate of brand stores based on data lakes according to claim 1, wherein in the step S1:
store information Store (poi): inputting information of a brand store to be collected, including specific information of the brand, specific location of the store, type of the location, place, floor where the store is located, number of the house, area, time of opening business and rent, from a store information management system;
collection position information Loc _ info (place): the method comprises the steps of including provincial and urban areas, longitude and latitude, business circles, city grades and city type basic position information;
collecting site information Place _ info (Place): the method comprises the following steps of including site addresses, site types, site ids, site areas, construction time, site floors, site average rent, site brand preference, passenger flow level, consumption level site label drawings, site business state distribution, site passenger group crowd drawings, site historical transaction information and site floor business state distribution, wherein the site types relate to shopping malls, office buildings and scenic spots, street pavements and districts, and the sites refer to specific shopping centers, office buildings, districts, building names and numbers;
collecting peripheral information Sur _ info (place) including transportation facilities and public facilities around a site;
the collected environment information Context comprises weather conditions, temperature, season, holidays, weeks and promotion activities;
acquiring Brand information Brand _ info (Brand) of a merchant, wherein the Brand information Brand _ info (Brand) comprises industries, enterprises, numbers of chain stores, brand positioning, target customer groups, product average price and competitive Brand;
the collected passenger Flow data Flow (poi) includes the number of people entering the store, the number of people passing the store, the collection time, and the attributes of the passenger group.
3. The method for predicting the store-in rate of brand stores based on data lake according to claim 1, wherein the passenger flow data in step S1 comprises: through passenger flow detection equipment comprising a fixed camera and mobile video shooting equipment, acquiring the number of people entering a store, the number of people passing the store and corresponding data acquisition time thereof by using a human body recognition and target tracking technology, and sending data to a back-end database; and collecting the passenger flow data of different shops of different brands together to form a lake of the passenger flow data of the shops.
4. The method for predicting the store-in rate of brand stores based on data lakes according to claim 1, wherein in the step S2: and (3) processing historical data by adopting the promotion environment information in the step (S1), wherein the promotion days are removed in a processing mode, or the increase rate of the promotion days relative to the daily store-in rate is calculated, and the store-in rate of the promotion days is reduced to the daily.
5. The method for predicting the store-in rate of brand stores based on data lake according to claim 1, wherein step S3 comprises: because the future environmental information is unknown, a prediction target and a sample of a designated dimension can be set according to a business target, a model is built according to different dimensions, the dimensions are divided according to a plurality of dimensions of a month and a working day and a working week, an average store-entering rate prediction target of a dimensional store (poi) is formed, and long-term and short-term prediction tasks are realized.
6. The data lake based store-entry rate prediction method for brand stores as claimed in claim 1,
the data structure characteristics after processing specifically include: after data are aggregated through specified dimensions, generating characteristics required by a model through characteristic conversion, characteristic coding and characteristic combination; the environmental information characteristics are not preserved after aggregation.
7. The method for predicting the store-in rate of a brand store based on a data lake according to claim 1, wherein the store-in rate modeling based on binomial distribution is used for constructing a maximum likelihood function to be solved as an optimization objective function, and the maximum likelihood function is as follows:
wherein, T i Number of total store-passing persons representing ith time period of brand store, N i Indicates the total number of store entries, X i Characteristic factor for showing influence on store-in rate of brand stores, store-in rate
w is a parameter, which needs to be estimated by a sample; pi (X) i ) Other forms are possible, depending on the particular problem; the time interval division canThe total amount is assumed to be N segments, namely N samples, by day or week or month; the problem of parameter w estimation is to find the optimal w to maximize L (w).
8. The method for predicting the store-in rate of brand stores based on data lakes according to claim 1, wherein in the step S3: all samples are divided into a training set, a verification set and a test set, the training set, the verification set and the test set are used for training and parameter adjustment of open source algorithms in machine learning and deep learning, and finally a model with the optimal performance of the test set is selected to retrain all samples to obtain a model to be on-line.
9. A data lake based store-entry rate prediction system for a brand store, comprising:
the data acquisition module is used for acquiring basic data required by model training, and the basic data comprises Store information Store (poi), location information Loc _ info (place), site information (place), peripheral information Sur _ info (place), merchant Brand information Brand _ info (Brand), environment information Context and passenger Flow data Flow (poi);
the data processing module is used for processing and cleaning the collected data, processing abnormal values and removing noise;
the model construction module is used for generating samples and characteristics required by the model, and further comprises a prediction target dimension determination unit which constructs targets and samples according to the set dimensions;
the model training module is used for inputting the training samples into the model to obtain a store-entering rate prediction model, and further comprises a sample dividing unit which is used for dividing a training set, a verification set and a test set, a model training parameter searching unit, an optimal model obtaining unit for various algorithms and a model training unit to be on-line, and is used for inputting all the samples into the optimal model to obtain a final model to be on-line;
and the online prediction module is used for evaluating the short-term and long-term future store-entering rate of the planned-operation position brand, inputting the characteristics required by the model and obtaining a store-entering rate prediction result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211548984.8A CN115860812A (en) | 2022-12-05 | 2022-12-05 | Brand store entrance rate prediction method and system based on data lake |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211548984.8A CN115860812A (en) | 2022-12-05 | 2022-12-05 | Brand store entrance rate prediction method and system based on data lake |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115860812A true CN115860812A (en) | 2023-03-28 |
Family
ID=85669901
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211548984.8A Pending CN115860812A (en) | 2022-12-05 | 2022-12-05 | Brand store entrance rate prediction method and system based on data lake |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115860812A (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108985865A (en) * | 2018-08-27 | 2018-12-11 | 广州联欣自动识别技术有限公司 | The customer data analysis method and system of intelligent shops |
| CN111242666A (en) * | 2018-11-29 | 2020-06-05 | Tcl集团股份有限公司 | Store site selection method based on big data analysis, storage medium and server |
| CN114386708A (en) * | 2022-01-19 | 2022-04-22 | 上海兴容信息技术有限公司 | Method and system for predicting door store passenger flow |
| CN115222123A (en) * | 2022-07-14 | 2022-10-21 | 中国建设银行股份有限公司 | Customer quantity prediction method, apparatus, device, medium, and program product |
| CN115239372A (en) * | 2022-06-29 | 2022-10-25 | 杭州路过网络有限公司 | Point location fragment passenger flow-based merchant brand-point location marketing effect estimation method |
| CN115239369A (en) * | 2022-06-29 | 2022-10-25 | 杭州路过网络有限公司 | Point location passenger flow volume prediction method based on point location segment passenger flow |
-
2022
- 2022-12-05 CN CN202211548984.8A patent/CN115860812A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108985865A (en) * | 2018-08-27 | 2018-12-11 | 广州联欣自动识别技术有限公司 | The customer data analysis method and system of intelligent shops |
| CN111242666A (en) * | 2018-11-29 | 2020-06-05 | Tcl集团股份有限公司 | Store site selection method based on big data analysis, storage medium and server |
| CN114386708A (en) * | 2022-01-19 | 2022-04-22 | 上海兴容信息技术有限公司 | Method and system for predicting door store passenger flow |
| CN115239372A (en) * | 2022-06-29 | 2022-10-25 | 杭州路过网络有限公司 | Point location fragment passenger flow-based merchant brand-point location marketing effect estimation method |
| CN115239369A (en) * | 2022-06-29 | 2022-10-25 | 杭州路过网络有限公司 | Point location passenger flow volume prediction method based on point location segment passenger flow |
| CN115222123A (en) * | 2022-07-14 | 2022-10-21 | 中国建设银行股份有限公司 | Customer quantity prediction method, apparatus, device, medium, and program product |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107844915B (en) | Automatic scheduling method of call center based on traffic prediction | |
| CN109214863B (en) | Method for predicting urban house demand based on express delivery data | |
| CN107423859B (en) | Combined mold building method and system | |
| AU2020102465A4 (en) | A method of predicting housing price using the method of combining multiple source data with mathematical model | |
| CN111582559B (en) | Arrival time estimation method and device | |
| Cho et al. | Relationship between value of open space and distance from housing locations within a community | |
| CN113379227A (en) | Industrial park data processing method and device, computer equipment and storage medium | |
| CN109741621B (en) | Parking space information processing method and system based on crowd sensing and computer device | |
| CN109166034A (en) | A kind of Risk Forecast Method and system | |
| CN113205383A (en) | Intelligent house property recommendation method and system | |
| Cullinan | A spatial microsimulation approach to estimating the total number and economic value of site visits in travel cost modelling | |
| CN115456695A (en) | Method, device, system and medium for analyzing shop address selection | |
| Assemi et al. | On-street parking occupancy inference based on payment transactions | |
| CN115034429A (en) | Urban taxi pick-up point recommendation method based on depth map neural network | |
| CN117273281A (en) | Intelligent travel data analysis method and system based on RFID technology | |
| JP7071948B2 (en) | Programs, equipment and methods for estimating the occupancy rate of real estate in the target area | |
| Shah et al. | A water demand prediction model for central indiana | |
| CN110309953B (en) | Urban security monitoring layout system and method adopting target mobility distribution prediction | |
| CN115860812A (en) | Brand store entrance rate prediction method and system based on data lake | |
| CN115239372A (en) | Point location fragment passenger flow-based merchant brand-point location marketing effect estimation method | |
| CN109767079A (en) | Region comparative approach is approached on a kind of boundary of energy internet business model assessment | |
| CN114387816B (en) | Mall traffic guiding system and method based on consumer behavior prediction analysis | |
| CN113902481B (en) | Rights and interests determining method, device, storage medium and apparatus | |
| Karunanithi et al. | A predictive analytics approach for forecasting bike rental demand | |
| CN110837903A (en) | Rail transit short-term passenger flow prediction method and system and electronic equipment |
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 |