CN109241227B - Spatiotemporal data prediction modeling method based on stacking integrated learning algorithm - Google Patents
Spatiotemporal data prediction modeling method based on stacking integrated learning algorithm Download PDFInfo
- Publication number
- CN109241227B CN109241227B CN201811017912.4A CN201811017912A CN109241227B CN 109241227 B CN109241227 B CN 109241227B CN 201811017912 A CN201811017912 A CN 201811017912A CN 109241227 B CN109241227 B CN 109241227B
- Authority
- CN
- China
- Prior art keywords
- data
- time
- space
- stacking
- model
- 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.)
- Active
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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/285—Selection of pattern recognition techniques, e.g. of classifiers in a multi-classifier system
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Data Mining & Analysis (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Operations Research (AREA)
- Entrepreneurship & Innovation (AREA)
- Quality & Reliability (AREA)
- General Business, Economics & Management (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
- Marketing (AREA)
- Tourism & Hospitality (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a spatiotemporal data prediction modeling method based on a stacking integrated learning algorithm, which can improve spatiotemporal data modeling efficiency and overall model training effect. The modeling method is based on mass data, adopts a stacking integrated learning method to realize data-driven space-time data prediction modeling, avoids the complex space-time data statistical modeling process in the past, and improves the efficiency of space-time data modeling; the stacking space-time data modeling technology gives consideration to the characteristics of processing time, space characteristics, dynamic characteristics and static characteristics, realizes secondary processing generation of the characteristics by a stacking method, and improves the training effect of the whole model; the invention discloses a stacking space-time prediction modeling technology, which adopts decision trees, GBDT, random forests and the like as base models, and has the following functions relative to a depth network model: less sample data, lower time complexity, non-black box of model results, and the like. Is suitable for popularization and application in the technical field of data processing.
Description
Technical Field
The invention relates to the technical field of data processing, in particular to a spatio-temporal data prediction modeling method based on a stacking integrated learning algorithm.
Background
Spatio-temporal data is data having both temporal and spatial dimensions, with more than 80% of the data in the real world being related to geographic location. As the world becomes instrumented and interrelated, spatiotemporal data is more prevalent and enriched than ever before, and the acquisition of complex patterns in spatiotemporal data by spatiotemporal data prediction techniques is also becoming more important and urgent for spatiotemporal data research applications.
The track of moving objects (e.g., taxis) recorded by GPS devices, social events (e.g., microblogs, crimes) with location markers and time stamps, and environmental monitoring are typical spatiotemporal data. These emerging spatiotemporal data also present new challenges and opportunities for data analysis research. On the one hand, data has spatial heterogeneity and autocorrelation, and spatial heterogeneity and spatial relationships (e.g., topological relationships, directional relationships, etc.) need to be processed. On the other hand, spatio-temporal data is dynamic in time, requiring explicit or implicit modeling of spatio-temporal autocorrelation and constraints to achieve good predictive performance.
Spatiotemporal sequence data is a very important type of spatiotemporal data. The study of spatio-temporal sequence data, which is generally considered as a collection of spatially correlated time sequences, has resulted in a number of approaches. Comprising the following steps: space-time autoregressive moving average (STARMA) model, to space-time dynamics methods, space-time regression statistical methods, to hybrid models, deep learning models, and the like.
The STARMA technology realizes the improvement of the time sequence ARMA model in the time-space data by adopting a linear mechanism that the time-space delay operator expresses that the time-space data is influenced by time and space, is a first time-space integrated modeling technology, and has successful application in a plurality of fields such as economy, climate, traffic, price prediction and the like. The STARMA requires sequence data to have stationarity, and the modeling process generally comprises three steps of model identification, parameter estimation and model inspection, and is complex.
The non-stationary spatio-temporal sequence mixing (Hybrid) modeling technique is a method of modeling for non-stationary spatio-temporal sequences. The core idea is that the technology such as regression or neural network extracts the large-scale space-time law of space-time data, and then models the small-scale variation of the space by STARMA and the like. In addition, methods such as hierarchical bayesian (Hierarchical Bayes Model), state space model, kalman filtering, etc. are also widely used for non-stationary spatio-temporal sequence predictive modeling. These modeling techniques rely in part on knowledge of the mechanism and the modeling process is still relatively complex.
In recent years, deep learning technology has been rapidly developed, and deep learning models in the field of spatio-temporal data prediction have also been proposed, including models such as ConvLSTM, deepST. The development of the technology greatly improves the performance of space-time data prediction in certain fields (such as weather rainfall). However, deep learning is generally dependent on large-scale available data, and has a great prospect in the fields of weather, traffic and the like. However, the depth network is often a 'black box', modeling time and space complexity are large, and massive training samples are needed, so that uncertainty expression capability in time data is weak.
Disclosure of Invention
The invention aims to solve the technical problem of providing a spatiotemporal data prediction modeling method based on a stacking integrated learning algorithm, which can improve the spatiotemporal data modeling efficiency and the overall model training effect.
The technical scheme adopted for solving the technical problems is as follows: the spatiotemporal data prediction modeling method based on the stacking integrated learning algorithm comprises the following steps:
A. extracting space-time source data in a period of historical time according to the needs of space-time data prediction tasks;
B. carrying out space-time data processing on the extracted space-time source data to obtain a dynamic characteristic data set in time, space or space-time dimension;
C. setting a time division point T 0 Dividing the dynamic characteristic Data set obtained in the step B into a first layer Data set Data 1 And a second layer Data set Data 2 First layer Data set Data 1 And a second layer Data set Data 2 The segmentation criteria of (2) are: if the time corresponding to the data in the dynamic characteristic data set is less than the time division point T 0 Dividing the Data into a first layer Data set Data 1 In the dynamic characteristic data set, the time corresponding to the data is longer than the time division point T 0 Dividing the Data into a second layer Data set Data 2 In (a) and (b);
D. randomly splitting data in a first layer of data set into n parts to obtain a data set { data } 11 ,data 12 ,…,data 1n }；
E. Data set { data } 11 ,data 12 ,…,data 1n Each subset data in } 1i (i∈[1,n]) Training a base model to obtain a base model set { base }, for training samples 1 ,basem 2 ,…,basem n }；
F. Predicting a second tier dataset Data using a base model 2 Results { pred 1 ,pred 2 ,…,pred n }；
G. Data set Data of the second layer 2 Is a static spatial feature of (1) and a prediction result { pred) of a base model 1 ,pred 2 ,…,pred n Fusion into a higher order feature data set NewData 2 ；
H. In NewData 2 Training a stacking model on the data set to obtain a stacking model;
I. the trained first layer { base } 1 ,basem 2 ,…,basem n And forming an integral model structure by the second layer model stack to obtain a space-time data prediction model.
Further, in the step B, the spatio-temporal data processing includes three basic data processing procedures of data anomaly identification, spatio-temporal data fusion and spatio-temporal dynamic characteristic data generation.
Further, in step C, the first layer Data set Data 1 Data volume and second layer Data set Data 2 The ratio of the data amounts of (2) is 7:3 or 6:4.
Further, in step E, training of each base model includes parameter optimization, model training, and data prediction.
Further, a random search algorithm or a Bayesian search algorithm is adopted in the parameter optimizing algorithm in the base model.
Further, the model training algorithm in the base model adopts a decision tree or random forest or GBDT.
Further, in step H, the training of the stacking model includes parameter optimization, learning training, and data prediction.
Further, a random search algorithm or a Bayesian search algorithm is adopted in the parameter optimizing algorithm in the stacking model.
Furthermore, the learning training algorithm in the stacking model selects logistic regression as a learning algorithm in a classification task, and linear regression is adopted as the learning algorithm in a regression task.
The invention has the beneficial effects that: the spatiotemporal data prediction modeling method based on the stacking integrated learning algorithm realizes data-driven spatiotemporal data prediction modeling by adopting the stacking integrated learning method on the basis of mass data, avoids the complex spatiotemporal data statistical modeling process in the past, and improves the efficiency of spatiotemporal data modeling; the stacking space-time data modeling technology gives consideration to the characteristics of processing time, space characteristics, dynamic characteristics and static characteristics, realizes secondary processing generation of the characteristics by a stacking method, and improves the training effect of the whole model; the invention discloses a stacking space-time prediction modeling technology, which adopts decision trees, GBDT, random forests and the like as base models, and has the following functions relative to a depth network model: less sample data, lower time complexity, non-black box of model results, and the like.
Detailed Description
The spatiotemporal data prediction modeling method based on the stacking integrated learning algorithm comprises the following steps:
A. extracting space-time source data in a period of historical time according to the needs of space-time data prediction tasks; such as the highest temperature of a pixel in the last 10 days, the total rainfall of the pixel in the last 10 days, the highest temperature in the range of 1 km around the pixel in the last 10 days, etc.;
B. carrying out space-time data processing on the extracted space-time source data to obtain a dynamic characteristic data set in time, space or space-time dimension;
C. setting a time division point T 0 Dividing the dynamic characteristic Data set obtained in the step B into a first layer Data set Data 1 And a second layer Data set Data 2 First layer Data set Data 1 And a second layer Data set Data 2 The segmentation criteria of (2) are: if the time corresponding to the data in the dynamic characteristic data set is less than the time division point T 0 Dividing the Data into a first layer Data set Data 1 In the dynamic characteristic data set, the time corresponding to the data is longer than the time division point T 0 Dividing the Data into a second layer Data set Data 2 In (a) and (b);
D. randomly splitting data in a first layer of data set into n parts to obtain a data set { data } 11 ,data 12 ,…,data 1n }；
E. Data set { data } 11 ,data 12 ,…,data 1n Each subset data in } 1i (i∈[1,n]) Training a base model to obtain a base model set { base }, for training samples 1 ,basem 2 ,…,basem n }；
F. Predicting a second tier dataset Data using a base model 2 Results { pred 1 ,pred 2 ,…,pred n -a }; the classification task is a predicted probability value, and the regression task is a predicted value;
G. data set Data of the second layer 2 Is a static spatial feature of (1) and a prediction result { pred) of a base model 1 ,pred 2 ,…,pred n Fusion into a higher order feature data set NewData 2 ；
H. In NewData 2 Training a stacking model on the data set to obtain a stacking model;
I. the trained first layer { base } 1 ,basem 2 ,…,basem n And forming an integral model structure by the second layer model stack to obtain a space-time data prediction model.
The spatiotemporal data prediction model realizes data-driven spatiotemporal data prediction modeling by adopting a stacking integrated learning method on the basis of massive data, avoids the complex spatiotemporal data statistical modeling process in the past, and improves the efficiency of spatiotemporal data modeling; the stacking space-time data modeling technology gives consideration to the characteristics of processing time, space characteristics, dynamic characteristics and static characteristics, realizes secondary processing generation of the characteristics by a stacking method, and improves the training effect of the whole model; the invention discloses a stacking space-time prediction modeling technology, which adopts decision trees, GBDT, random forests and the like as base models, and has the following functions relative to a depth network model: less sample data, lower time complexity, non-black box of model results, and the like.
In order to make the prediction of the finally established prediction model more accurate, in the step B, the space-time data processing comprises data anomaly identification, space-time data fusion and space-time dynamicsThe state characteristic data generates three basic data processing procedures. In step C, the first layer Data set Data 1 Data volume and second layer Data set Data 2 The ratio of the data amounts of (2) is 7:3 or 6:4.
In addition, in step E, training of each base model includes parameter optimization, model training, and data prediction.
The optimization algorithm of the model mainly provides parameter searching and selection in the training process, and is characterized by massive and high-dimensional characteristics of space-time data. And the parameter optimizing algorithm in the base model adopts a random searching algorithm and a Bayesian searching algorithm to conduct model parameter searching optimization.
In view of the complexity and uncertainty of the space-time data, the algorithm adopted by the base model layer should have fast efficiency and nonlinear expression capability. The model training algorithm in the base model adopts decision trees, random forests, GBDT and the like.
In step H, training of the stacking model includes parameter optimization, learning training, and data prediction.
The optimization algorithm of the model mainly provides parameter searching and selection in the training process, and is characterized by massive and high-dimensional characteristics of space-time data. And the parameter optimizing algorithm in the stacking model adopts a random searching algorithm and a Bayesian searching algorithm to perform model parameter searching optimization.
Different algorithms are selected according to learning tasks, the learning training algorithm in the stacking model selects logistic regression as the learning algorithm in the classification tasks, and linear regression is adopted as the learning algorithm in the regression tasks.
Claims (8)
1. The spatio-temporal data prediction modeling method based on the stacking integrated learning algorithm is characterized by comprising the following steps of:
A. extracting space-time source data in a period of historical time according to the needs of space-time data prediction tasks;
B. carrying out space-time data processing on the extracted space-time source data to obtain a dynamic characteristic data set in time, space or space-time dimension; the space-time data processing comprises three basic data processing processes of data anomaly identification, space-time data fusion and space-time dynamic characteristic data generation;
C. setting a time division point T 0 Dividing the dynamic characteristic Data set obtained in the step B into a first layer Data set Data 1 And a second layer Data set Data 2 First layer Data set Data 1 And a second layer Data set Data 2 The segmentation criteria of (2) are: if the time corresponding to the data in the dynamic characteristic data set is less than the time division point T 0 Dividing the Data into a first layer Data set Data 1 In the dynamic characteristic data set, the time corresponding to the data is longer than the time division point T 0 Dividing the Data into a second layer Data set Data 2 In (a) and (b);
D. randomly splitting data in a first layer of data set into n parts to obtain a data set { data } 11 ,data 12 ,…,data 1n }；
E. Data set { data } 11 ,data 12 ,…,data 1n Each subset data in } 1i (i∈[1,n]) Training a base model to obtain a base model set { base }, for training samples 1 ,basem 2 ,…,basem n }；
F. Predicting a second tier dataset Data using a base model 2 Results { pred 1 ,pred 2 ,…,pred n }；
G. Data set Data of the second layer 2 Is a static spatial feature of (1) and a prediction result { pred) of a base model 1 ,pred 2 ,…,pred n Fusion into a higher order feature data set NewData 2 ；
H. In NewData 2 Training a stacking model on the data set to obtain a stacking model;
I. the trained first layer { base } 1 ,basem 2 ,…,basem n And forming an integral model structure by the second layer model stack to obtain a space-time data prediction model.
2. The spatiotemporal data prediction modeling method based on a stacking integrated learning algorithm of claim 1,the method is characterized in that: in step C, the first layer Data set Data 1 Data volume and second layer Data set Data 2 The ratio of the data amounts of (2) is 7:3 or 6:4.
3. The spatio-temporal data prediction modeling method based on a stacking ensemble learning algorithm as claimed in claim 2, wherein: in step E, training of each base model includes parameter optimization, model training and data prediction.
4. The spatio-temporal data prediction modeling method based on a stacking ensemble learning algorithm as claimed in claim 3, wherein: the parameter optimizing algorithm in the base model adopts a random searching algorithm or a Bayesian searching algorithm.
5. The spatio-temporal data prediction modeling method based on the stacking integrated learning algorithm according to claim 4, wherein: the model training algorithm in the base model adopts decision trees or random forests or GBDT.
6. The spatio-temporal data prediction modeling method based on the stacking integrated learning algorithm according to claim 5, wherein: in step H, training of the stacking model includes parameter optimization, learning training, and data prediction.
7. The spatio-temporal data prediction modeling method based on a stacking ensemble learning algorithm as claimed in claim 6, wherein: the parameter optimizing algorithm in the stacking model adopts a random searching algorithm or a Bayesian searching algorithm.
8. The spatio-temporal data prediction modeling method based on a stacking ensemble learning algorithm as claimed in claim 7, wherein: the learning training algorithm in the stacking model selects logistic regression as a learning algorithm in a classification task, and adopts linear regression as a learning algorithm in a regression task.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811017912.4A CN109241227B (en) | 2018-09-03 | 2018-09-03 | Spatiotemporal data prediction modeling method based on stacking integrated learning algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811017912.4A CN109241227B (en) | 2018-09-03 | 2018-09-03 | Spatiotemporal data prediction modeling method based on stacking integrated learning algorithm |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109241227A CN109241227A (en) | 2019-01-18 |
| CN109241227B true CN109241227B (en) | 2023-05-30 |
Family
ID=65059982
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811017912.4A Active CN109241227B (en) | 2018-09-03 | 2018-09-03 | Spatiotemporal data prediction modeling method based on stacking integrated learning algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109241227B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110347760B (en) * | 2019-05-30 | 2021-07-09 | 中国地质大学（武汉） | Data analysis method for lost crowd space-time positioning service |
| CN110245609A (en) * | 2019-06-13 | 2019-09-17 | 深圳力维智联技术有限公司 | Pedestrian track generation method, device and readable storage medium storing program for executing |
| CN111475744B (en) * | 2020-04-03 | 2022-06-14 | 南京理工大学紫金学院 | Personalized position recommendation method based on ensemble learning |
| CN112756759B (en) * | 2021-01-11 | 2022-04-08 | 上海智能制造功能平台有限公司 | Spot welding robot workstation fault judgment method |
| CN112904157A (en) * | 2021-01-19 | 2021-06-04 | 重庆邮电大学 | Fault arc detection method based on integrated machine learning |
| CN112985574B (en) * | 2021-02-26 | 2022-02-01 | 电子科技大学 | High-precision classification identification method for optical fiber distributed acoustic sensing signals based on model fusion |
| CN113722288A (en) * | 2021-07-27 | 2021-11-30 | 张博 | Modeling method for time-space data statistics |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103257921B (en) * | 2013-04-16 | 2015-07-22 | 西安电子科技大学 | Improved random forest algorithm based system and method for software fault prediction |
| CN107301221A (en) * | 2017-06-16 | 2017-10-27 | 华南理工大学 | A kind of data digging method of multiple features dimension heap fusion |
| CN107507038B (en) * | 2017-09-01 | 2021-03-19 | 美林数据技术股份有限公司 | Electricity charge sensitive user analysis method based on stacking and bagging algorithms |
| CN108090607A (en) * | 2017-12-13 | 2018-05-29 | 中山大学 | A kind of social media user's ascribed characteristics of population Forecasting Methodology based on the fusion of multi-model storehouse |
| CN108375808A (en) * | 2018-03-12 | 2018-08-07 | 南京恩瑞特实业有限公司 | Dense fog forecasting procedures of the NRIET based on machine learning |
-
2018
- 2018-09-03 CN CN201811017912.4A patent/CN109241227B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN109241227A (en) | 2019-01-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109241227B (en) | Spatiotemporal data prediction modeling method based on stacking integrated learning algorithm | |
| CN110245981B (en) | Crowd type identification method based on mobile phone signaling data | |
| CN109034448B (en) | Trajectory prediction method based on vehicle trajectory semantic analysis and deep belief network | |
| CN108898829B (en) | Dynamic short-time traffic flow prediction system aiming at non-difference division and data sparseness | |
| Okawa et al. | Deep mixture point processes: Spatio-temporal event prediction with rich contextual information | |
| CN109145175B (en) | Spatiotemporal data prediction method based on stacking integrated learning algorithm | |
| CN106528874B (en) | The CLR multi-tag data classification method of big data platform is calculated based on Spark memory | |
| CN111932026B (en) | Urban traffic pattern mining method based on data fusion and knowledge graph embedding | |
| CN113157800B (en) | Identification method for discovering dynamic target in air in real time | |
| CN113159364A (en) | Passenger flow prediction method and system for large-scale traffic station | |
| CN105493109A (en) | Air quality inference using multiple data sources | |
| Li et al. | A DBN-based deep neural network model with multitask learning for online air quality prediction | |
| CN110335507A (en) | Flight operation situation law analytical method based on blank pipe track big data | |
| CN105046714A (en) | Unsupervised image segmentation method based on super pixels and target discovering mechanism | |
| CN115238197B (en) | Expert thinking model-based domain business auxiliary analysis method | |
| CN111242352A (en) | Parking aggregation effect prediction method based on vehicle track | |
| Irfan et al. | Performance analysis of machine learning techniques for wind speed prediction | |
| Wang et al. | R2-trans: Fine-grained visual categorization with redundancy reduction | |
| Viswambari et al. | Data mining techniques to predict weather: a survey | |
| CN112101132A (en) | Traffic condition prediction method based on graph embedding model and metric learning | |
| Aydın | Classification of the fire station requirement with using machine learning algorithms | |
| Zhang | Remote sensing data processing of urban land using based on artificial neural network | |
| Vardhan et al. | Density based clustering technique on crop yield prediction | |
| Sukhija et al. | Spatial and temporal trends reveal: Hotspot identification of crimes using machine learning approach | |
| Yang et al. | A data-driven method for flight time estimation based on air traffic pattern identification and prediction |
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 | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20230113 Address after: 610000 1, 3, 1, 366 north section of lakeside road, Tianfu New District, Chengdu, Sichuan Applicant after: Chengdu Cap Data Service Co.,Ltd. Address before: Room 503, 5th floor, unit 1, building 12, 333 Taihe 2nd Street, high tech Zone, Chengdu, Sichuan 610041 Applicant before: SICHUAN JIALIAN ZHONGHE ENTERPRISE MANAGEMENT CONSULTATION Co.,Ltd. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |