CN106611227B - Method and system for predicting dynamic change curve of flow of interest point - Google Patents
Method and system for predicting dynamic change curve of flow of interest point Download PDFInfo
- Publication number
- CN106611227B CN106611227B CN201510698197.5A CN201510698197A CN106611227B CN 106611227 B CN106611227 B CN 106611227B CN 201510698197 A CN201510698197 A CN 201510698197A CN 106611227 B CN106611227 B CN 106611227B
- Authority
- CN
- China
- Prior art keywords
- day
- flow
- prediction
- orthogonal basis
- training
- 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
Images
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
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- Strategic Management (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- General Physics & Mathematics (AREA)
- Development Economics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Game Theory and Decision Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Engineering & Computer Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
A method for predicting a flow dynamic change curve of an interest point comprises the following steps: extracting the flow information of the interest points from a historical database, and performing time-sharing statistics on the flow information of the interest points in one day to obtain a flow change curve of each day in a certain time period; generating a training sample set based on a flow change curve of each day in a certain time period by combining a method based on standard orthogonal basis decomposition; training an SVM model by utilizing a training sample set for the decomposition coefficient under each orthogonal basis to obtain a corresponding SVM regression prediction function; performing orthogonal basis decomposition coefficient prediction by using the obtained SVM regression prediction function; and predicting the flow dynamic change curve of the interest point by using the orthogonal basis decomposition coefficient. The dynamic flow curve prediction method can effectively improve the prediction precision and meet the requirement of dynamic prediction of the flow of the POI point of the actual map.
Description
Technical Field
The invention belongs to the field of information technology and automatic control, and particularly relates to a method for predicting a flow change curve of a future day according to a daily flow change curve counted by historical data of POI points.
Background
With the popularization of map applications, the requirement for inquiring points of interest (POI for short in english) in a map is higher and higher, and it is desirable to not only inquire the position of a certain POI point, but also inquire various attribute information such as personnel traffic, scenic spot evaluation, surrounding weather and the like. Various rich POI attribute information is the basis for improving the use performance of the navigation map and is the key for improving the satisfaction degree of a user on the map searching performance. Traffic information is one of the important attributes of point of interest POI.
However, in the process of predicting the traffic of a POI point of a certain scenic spot, the existing prediction methods mostly adopt methods such as regression to predict the total traffic information of a certain day, and a method for predicting dynamic change information of traffic of the whole day is lacked, so that the method is difficult to meet the query requirement of a user on the dynamic change situation of the day, and how to accurately predict the dynamic change situation of the traffic of the POI point of a map in the day has important significance in aspects such as trip planning and scenic spot management of the user.
Disclosure of Invention
In order to overcome the defect that the prediction accuracy is difficult to meet the requirement of practical application because the existing prediction method does not consider the correlation existing in the flow among different moments in the flow prediction process and only carries out static prediction, the method for predicting the dynamic change curve of the flow of the POI point can be provided, and the flow information of each moment in a day can be predicted.
In order to achieve the above object, the present invention provides a method for predicting a dynamic change curve of a point of interest flow, including:
step 1), extracting flow information of interest points from a historical database, and performing time-sharing statistics on the flow information of the interest points in one day to obtain a flow change curve of each day in a certain time period;
step 2), generating a training sample set based on the flow change curve obtained in the step 1) every day in a certain time period by combining a method based on standard orthogonal basis decomposition; the input feature set of the training sample set comprises orthogonal basis decomposition coefficients of a plurality of days before the day to be predicted, and the output target set comprises the orthogonal basis decomposition coefficients of the day to be predicted;
step 3), training an SVM model by using the training sample set obtained in the step 2) for the decomposition coefficient under each orthogonal basis to obtain a corresponding SVM regression prediction function;
step 4), carrying out orthogonal basis decomposition coefficient prediction by using the SVM regression prediction function obtained in the step 3);
and 5) predicting a flow dynamic change curve of the interest point by using the orthogonal basis decomposition coefficient obtained by prediction in the step 4).
In the above technical solution, the step 2) further includes:
step 2-1), expanding the daily flow rate change curve of the historical data obtained in the step 1) by adopting a method based on standard orthogonal basis decomposition to obtain an orthogonal basis decomposition coefficient β of the daily flow rate curveik;
The standard orthogonal base adopts a normalized trigonometric function set, the number of the selected standard orthogonal base is L, and the standard orthogonal base is in the form of:
the coefficients after expansion are:
wherein: vijThe statistics result at a certain time point, i represents the number of days, j represents the j statistic of the ith day, and j is 1,2, … and N;the value of the k-th basis function of phi (x) at point j βikThe expansion coefficient of the flow change curve of the ith day on the kth basis function is shown; 1,2,3, …, k 1,2,3, …, L;
step 2-2), generating a training sample set, comprising: taking the coefficients of the former Z days under a certain orthogonal base as input to form an input sample xm={βi-Z,k,βi-Z+1,k,…,βi-1,kWhere M is 1,2, …, and M is the total number of training samples; taking the decomposition coefficient under the orthogonal base at the Z +1 th day as an output ym={βi,k};
For M training samples, the training sample set is formed as follows:
an input feature set: x ═ X1,x2,…,xM];
Outputting a target set: f. of=[y1,y2,…,yM]。
In the above technical solution, the step 3) further includes:
step 3-1), solving an optimization problem by utilizing a quadratic programming algorithm:
s.t.
wherein, for a given parameter value, αi *、αiParameters found for training, yiFor the target output value of the training set, K (x)i,xj) Is a radial basis function kernel of the form:
xiis the input characteristic vector, gamma is the width parameter of the Gaussian kernel function;
step 3-2), after training is finished, establishing an SVM regression prediction function as follows:
where b is the trained threshold, αi *、αiParameters found for training; x is a given sample feature vector to be predicted;is a predicted value for x.
In the above technical solution, the step 4) further includes:
using the decomposition coefficient of the dynamic flow change curve of Z days before the current day under the standard orthogonal basis as an input feature vector, and utilizing the SVM loop established under each orthogonal basisAnd predicting the coefficients under the orthogonal basis of the Z +1 th day in the future by the prediction function, and recording the prediction results of all the coefficients as:
in the above technical solution, the step 5) further includes:
step 5-1), predicting the flow value of each time point in the 11 th day by using the following formula in combination with the orthogonal basis function decomposition coefficient obtained by prediction in the step 4);
in the above formula VikFlow values representing the ith time point on day i;
and 5-2) connecting the prediction results of different time points in the day to form a flow dynamic change prediction curve of the day.
The invention also provides a system for predicting the flow dynamic change curve of the interest point, which comprises the following steps:
the time-sharing statistic module extracts the flow information of the interest points from the historical database, and carries out time-sharing statistics on the flow information of the interest points in one day to obtain a flow change curve of each day in a certain time period;
the training sample set generating module is used for generating a training sample set by combining a method based on standard orthogonal basis decomposition based on a flow change curve obtained by the time-sharing statistic module every day in a certain time period; the input feature set of the training sample set comprises orthogonal basis decomposition coefficients of a plurality of days before the day to be predicted, and the output target set comprises the orthogonal basis decomposition coefficients of the day to be predicted;
an SVM regression prediction module, which trains an SVM model by using the training sample set obtained by the training sample set generation module for the decomposition coefficient under each orthogonal basis to obtain a corresponding SVM regression prediction function;
an orthogonal basis decomposition coefficient prediction module which performs orthogonal basis decomposition coefficient prediction by using an SVM regression prediction function obtained by the SVM regression prediction module;
and the interest point flow dynamic change curve generation module predicts the interest point flow dynamic change curve by using the orthogonal base decomposition coefficient obtained by the orthogonal base decomposition coefficient prediction module.
The invention has the advantages that:
the invention overcomes the defect that the prediction accuracy is difficult to meet the actual application requirement because the existing prediction method only carries out static prediction without considering the correlation existing in the flow between different moments in the flow prediction process, can realize the prediction of the dynamic change of the all-day flow of the POI point, considers the dynamic characteristic of the flow change and improves the prediction accuracy.
Drawings
Fig. 1 is a flowchart of a method for predicting a POI point traffic dynamic change curve based on a process support vector machine.
Detailed Description
The method for predicting the POI point flow dynamic change curve based on the process support vector machine is realized by prediction algorithm control software depending on related hardware equipment such as a database system, a prediction algorithm server, a user client and the like.
The following describes in detail the steps of the method for predicting the POI point traffic dynamic change curve based on the process support vector machine proposed in the present invention with reference to fig. 1:
step 1), extracting flow information of POI points from a historical database, and performing time-sharing statistics on the flow information of the POI points in one day to obtain a flow change curve of each day in a certain time period;
the method specifically comprises the following steps: and inquiring according to the time range from the historical database, and carrying out statistics at regular intervals (such as 5 minutes) to obtain the flow conditions of different time points in one day, thereby obtaining the flow change curve of each day. Wherein, the statistical result at a certain time point is recorded as VijI denotes the day number, j denotes the j-th statistic on day i, j is 1,2, …, N.
And 2) generating a training sample set based on the daily flow change curve obtained in the step 1).
The method specifically comprises the following steps:
step 2-1), expanding the daily flow rate change curve of the historical data obtained in the step 1) by adopting a method based on standard orthogonal basis decomposition to obtain an orthogonal basis decomposition coefficient β of the daily flow rate curveik;
The standard orthogonal base adopts a normalized trigonometric function set, the number of the selected standard orthogonal base is L, and the standard orthogonal base is in the form of:
the coefficients after expansion are:
wherein:the value of the k-th basis function of phi (x) at point j βikThe expansion coefficient of the flow change curve of the ith day on the kth basis function is shown; i is 1,2,3, …, k is 1,2,3, …, L.
Step 2-2), generating a training sample set, comprising: taking the coefficients under a certain orthogonal base in the previous 10 days as input to form an input sample xm={βi-10,k,βi-9,k,…,βi-1,kWhere M is 1,2, …, and M is the total number of training samples; the decomposition coefficient at the orthogonal basis at day 11 is taken as the output ym={βi,k}。
For M training samples, the training sample set is formed as follows:
an input feature set: x ═ X1,x2,…,xM];
Outputting a target set: f ═ y1,y2,…,yM]。
The number of days for which samples are input is not limited to the 10 days mentioned here, but may be other values.
Step 3), training an SVM model by using the training sample set obtained in the step 2) for the coefficient under each orthogonal basis to obtain a corresponding SVM regression prediction function;
the method specifically comprises the following steps:
step 3-1), solving an optimization problem by utilizing a quadratic programming algorithm:
s.t.
wherein, for a given parameter value, αi *、αiParameters found for training, yiFor the target output value of the training set, K (x)i,xj) Is a radial basis function kernel of the form:
xiis the input characteristic vector, gamma is the width parameter of the Gaussian kernel function;
step 3-2), after training is finished, establishing an SVM regression prediction function as follows:
where b is the trained threshold, αi *、αiParameters found for training; x is a given sample feature vector to be predicted;is a predicted value for x;
step 4), carrying out orthogonal basis decomposition coefficient prediction by using the SVM regression prediction function obtained in the step 3);
the method specifically comprises the following steps: with dynamic traffic variation 10 days before the current dayDecomposing coefficients of the curve under the standard orthogonal basis are used as input feature vectors, an SVM regression prediction function established under each orthogonal basis is used for predicting the coefficients under the orthogonal basis on the 11 th day in the future, and prediction results of all the coefficients are recorded as:
step 5), predicting a POI point flow dynamic change curve by using the orthogonal basis decomposition coefficient obtained by prediction in the step 4);
the method specifically comprises the following steps:
step 5-1), predicting the flow value of each time point in the 11 th day by using the following formula in combination with the orthogonal basis function decomposition coefficient obtained by prediction in the step 4);
in the above formula VikFlow values representing the ith time point on day i;
and 5-2) connecting the prediction results of different time points in the day to form a flow dynamic change prediction curve of the day.
Finally, it should be noted that the above embodiments are only used for illustrating the technical solutions of the present invention and are not limited. Although the present invention has been described in detail with reference to the embodiments, it will be understood by those skilled in the art that various changes may be made and equivalents may be substituted without departing from the spirit and scope of the invention as defined in the appended claims.
Claims (2)
1. A method for predicting a flow dynamic change curve of an interest point comprises the following steps:
step 1), extracting flow information of interest points from a historical database, and performing time-sharing statistics on the flow information of the interest points in one day to obtain a flow change curve of each day in a certain time period;
step 2), generating a training sample set based on the flow change curve obtained in the step 1) every day in a certain time period by combining a method based on standard orthogonal basis decomposition; the input feature set of the training sample set comprises orthogonal basis decomposition coefficients of a plurality of days before the day to be predicted, and the output target set comprises the orthogonal basis decomposition coefficients of the day to be predicted;
step 3), training an SVM model by using the training sample set obtained in the step 2) for the decomposition coefficient under each orthogonal basis to obtain a corresponding SVM regression prediction function;
step 4), carrying out orthogonal basis decomposition coefficient prediction by using the SVM regression prediction function obtained in the step 3);
step 5), predicting a flow dynamic change curve of the interest point by using the orthogonal basis decomposition coefficient obtained by prediction in the step 4);
the step 2) further comprises the following steps:
step 2-1), expanding the daily flow rate change curve of the historical data obtained in the step 1) by adopting a method based on standard orthogonal basis decomposition to obtain an orthogonal basis decomposition coefficient β of the daily flow rate curveik;
The standard orthogonal base adopts a normalized trigonometric function set, the number of the selected standard orthogonal base is L, and the standard orthogonal base is in the form of:
the coefficients after expansion are:
wherein: vijThe statistics result at a certain time point, i represents the number of days, j represents the j statistic of the ith day, and j is 1,2, … and N;the value of the k-th basis function of phi (x) at point j βikThe expansion coefficient of the flow change curve of the ith day on the kth basis function is shown; 1,2,3, …, k 1,2,3, …, L;
step 2-2), generating a training sample set, comprising: taking the coefficients of the former Z days under a certain orthogonal base as input to form an input sample xm={β1,k,β2,k,…,βZ,kWhere M is 1,2, …, and M is the total number of training samples; the decomposition coefficient at the orthogonal basis at the Z +1 th day is taken as an output, ym={βZ+1,k};
For M training samples, the training sample set is formed as follows:
an input feature set: x ═ X1,x2,…,xM];
Outputting a target set: f ═ y1,y2,…,yM];
The step 3) further comprises the following steps:
step 3-1), solving an optimization problem by utilizing a quadratic programming algorithm:
s.t.
wherein, for a given parameter value, αi *、αiParameters found for training, yiFor the target output value of the training set, K (x)i,xj) Is a radial basis function kernel of the form:
xiis the input characteristic vector, gamma is the width parameter of the Gaussian kernel function;
step 3-2), after training is finished, establishing an SVM regression prediction function as follows:
where b is the trained threshold, αi *、αiParameters found for training; x is a given sample feature vector to be predicted;is a predicted value for x;
the step 4) further comprises the following steps:
using the decomposition coefficient of the dynamic flow change curve of Z days before the ith day under the standard orthogonal basis as an input feature vector, and utilizing the SVM regression prediction function established under each orthogonal basisPredicting coefficients under orthogonal bases of the i-th day in the future, and recording prediction results of all the coefficients as:
the step 5) further comprises the following steps:
step 5-1), predicting the flow value of each time point in the ith day by using the following formula in combination with the orthogonal basis function decomposition coefficient obtained by prediction in the step 4);
in the above formula VijFlow values representing the ith time point on day i;
and 5-2) connecting the prediction results of different time points in the day to form a flow dynamic change prediction curve of the day.
2. A system for predicting a point of interest flow dynamics curve, comprising:
the time-sharing statistic module extracts the flow information of the interest points from the historical database, and carries out time-sharing statistics on the flow information of the interest points in one day to obtain a flow change curve of each day in a certain time period;
the training sample set generating module is used for generating a training sample set by combining a method based on standard orthogonal basis decomposition based on a flow change curve obtained by the time-sharing statistic module every day in a certain time period; the input feature set of the training sample set comprises orthogonal basis decomposition coefficients of a plurality of days before the day to be predicted, and the output target set comprises the orthogonal basis decomposition coefficients of the day to be predicted;
the method specifically comprises the following steps:
step 2-1), expanding the daily flow change curve of the historical data obtained by the time division statistical module by adopting a method based on standard orthogonal basis decomposition to obtain an orthogonal basis decomposition coefficient β of the daily flow curveik;
The standard orthogonal base adopts a normalized trigonometric function set, the number of the selected standard orthogonal base is L, and the standard orthogonal base is in the form of:
the coefficients after expansion are:
wherein: vijThe statistics result at a certain time point, i represents the number of days, j represents the j statistic of the ith day, and j is 1,2, … and N;the value of the k-th basis function of phi (x) at point j βikThe expansion coefficient of the flow change curve of the ith day on the kth basis function is shown; 1,2,3, …, k 1,2,3, …, L;
step 2-2), generating a training sample set, comprising: taking the coefficients of the first Z days of the ith day under a certain orthogonal basis as input to form an input sample xm={βi-Z,k,βi-Z+1,k,…,βi-1,kWhere M is 1,2, …, and M is the total number of training samples; decomposition at day i under this orthogonal basisCoefficient being output ym={βi,k};
For M training samples, the training sample set is formed as follows:
an input feature set: x ═ X1,x2,…,xM];
Outputting a target set: f ═ y1,y2,…,yM];
An SVM regression prediction module, which trains an SVM model by using the training sample set obtained by the training sample set generation module for the decomposition coefficient under each orthogonal basis to obtain a corresponding SVM regression prediction function; the method specifically comprises the following steps:
step 3-1), solving an optimization problem by utilizing a quadratic programming algorithm:
s.t.
wherein, for a given parameter value, αi *、αiParameters found for training, yiFor the target output value of the training set, K (x)i,xj) Is a radial basis function kernel of the form:
xiis the input characteristic vector, gamma is the width parameter of the Gaussian kernel function;
step 3-2), after training is finished, establishing an SVM regression prediction function as follows:
where b is the trained threshold, αi *、αiParameters derived for training(ii) a x is a given sample feature vector to be predicted;is a predicted value for x;
an orthogonal basis decomposition coefficient prediction module which performs orthogonal basis decomposition coefficient prediction by using an SVM regression prediction function obtained by the SVM regression prediction module; the method specifically comprises the following steps:
using the decomposition coefficient of the dynamic flow change curve of Z days before the ith day under the standard orthogonal basis as an input feature vector, and utilizing the SVM regression prediction function established under each orthogonal basisAnd predicting coefficients under the orthogonal basis of the day i, and recording the prediction results of all the coefficients as:
the interest point flow dynamic change curve generation module predicts an interest point flow dynamic change curve by using the orthogonal base decomposition coefficient obtained by the orthogonal base decomposition coefficient prediction module; the method specifically comprises the following steps:
predicting the flow value of each time point in the ith day by using the following formula in combination with the orthogonal basis function decomposition coefficient obtained by prediction;
in the above formula VijFlow values representing the ith time point on day i;
and 5-2) connecting the prediction results of different time points in the day to form a flow dynamic change prediction curve of the day.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510698197.5A CN106611227B (en) | 2015-10-23 | 2015-10-23 | Method and system for predicting dynamic change curve of flow of interest point |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510698197.5A CN106611227B (en) | 2015-10-23 | 2015-10-23 | Method and system for predicting dynamic change curve of flow of interest point |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106611227A CN106611227A (en) | 2017-05-03 |
CN106611227B true CN106611227B (en) | 2020-08-11 |
Family
ID=58612754
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510698197.5A Active CN106611227B (en) | 2015-10-23 | 2015-10-23 | Method and system for predicting dynamic change curve of flow of interest point |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106611227B (en) |
Families Citing this family (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112101682B (en) * | 2020-09-25 | 2024-04-09 | 北京百度网讯科技有限公司 | Traffic pattern prediction method, traffic pattern prediction device, server and readable medium |
CN112419729B (en) * | 2020-11-23 | 2022-04-15 | 北京嘀嘀无限科技发展有限公司 | Method and system for determining decision strategy according to predicted traffic flow |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103310287A (en) * | 2013-07-02 | 2013-09-18 | 北京航空航天大学 | Rail transit passenger flow predicting method for predicting passenger travel probability and based on support vector machine (SVM) |
CN103700255A (en) * | 2013-12-30 | 2014-04-02 | 复旦大学 | Time and space related data mining-based traffic flow prediction method |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8315786B2 (en) * | 2008-06-27 | 2012-11-20 | Microsoft Corporation | Local decision policies about the sharing of sensed data that enhance privacy and lower communication costs for services that aggregate data from personal devices |
-
2015
- 2015-10-23 CN CN201510698197.5A patent/CN106611227B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103310287A (en) * | 2013-07-02 | 2013-09-18 | 北京航空航天大学 | Rail transit passenger flow predicting method for predicting passenger travel probability and based on support vector machine (SVM) |
CN103700255A (en) * | 2013-12-30 | 2014-04-02 | 复旦大学 | Time and space related data mining-based traffic flow prediction method |
Non-Patent Citations (2)
Title |
---|
余向洋 等.基于EMD的景区客流波动特征及其组合预测--以黄山风景区为例.《地理科学进展》.2012,第31卷(第10期),1353-1359. * |
贾勇兵.基于SVM的交通流短时预测方法研究.《CNKI优秀硕士学位论文全文库》.2012,正文第13-14页,19页,34-42页,52-53页. * |
Also Published As
Publication number | Publication date |
---|---|
CN106611227A (en) | 2017-05-03 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110928993B (en) | User position prediction method and system based on deep cyclic neural network | |
CN104485665A (en) | Dynamic probabilistic power flow (PPF) calculating method considering wind speed predication error temporal-spatial coherence | |
CN106850289B (en) | Service combination method combining Gaussian process and reinforcement learning | |
CN112001422B (en) | Image mark estimation method based on deep Bayesian learning | |
WO2019169594A1 (en) | Methods and apparatus to generate three-dimensional (3d) model for 3d scene reconstruction | |
CN112910711A (en) | Wireless service flow prediction method, device and medium based on self-attention convolutional network | |
CN116776208B (en) | Training method of seismic wave classification model, seismic wave selecting method, equipment and medium | |
WO2023201938A1 (en) | Missing trajectory filling method and system | |
CN112258557B (en) | Visual tracking method based on space attention feature aggregation | |
CN105895089A (en) | Speech recognition method and device | |
CN106611227B (en) | Method and system for predicting dynamic change curve of flow of interest point | |
CN111325376A (en) | Wind speed prediction method and device | |
CN115965177A (en) | Improved autoregressive error compensation wind power prediction method based on attention mechanism | |
CN109254321A (en) | Quick Bayes's Modal Parameters Identification under a kind of seismic stimulation | |
CN112948101B (en) | 5G edge equipment scheduling method for rapid federal learning | |
US10169508B2 (en) | Efficient deployment of table lookup (TLU) in an enterprise-level scalable circuit simulation architecture | |
Bosma et al. | Estimating solar and wind power production using computer vision deep learning techniques on weather maps | |
CN115032682A (en) | Multi-station seismic source parameter estimation method based on graph theory | |
Tomaras | Crowdsourcing techniques for smart urban mobility | |
CN110543896B (en) | Heterogeneous crowdsourcing fingerprint labeling method based on semi-supervised naive Bayes | |
CN114117239A (en) | House resource pushing method, device and equipment | |
KR102390005B1 (en) | Learning method, learning device for generating high-resolution weather data and testing method, testing device using the same | |
CN109190697A (en) | A kind of hotel's picture classification judges system and method | |
Yuan et al. | Intelligent identification of ocean parameters based on RBF neural networks | |
JP7473837B2 (en) | Deterioration estimation device and deterioration estimation method |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |