CN111353523A - Method for classifying railway customers - Google Patents
Method for classifying railway customers Download PDFInfo
- Publication number
- CN111353523A CN111353523A CN201911343327.8A CN201911343327A CN111353523A CN 111353523 A CN111353523 A CN 111353523A CN 201911343327 A CN201911343327 A CN 201911343327A CN 111353523 A CN111353523 A CN 111353523A
- Authority
- CN
- China
- Prior art keywords
- railway
- model
- time sequence
- prophet
- client
- 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 40
- 238000012545 processing Methods 0.000 claims abstract description 18
- 230000008569 process Effects 0.000 claims abstract description 14
- 238000012549 training Methods 0.000 claims abstract description 13
- 230000015654 memory Effects 0.000 claims abstract description 3
- 238000010276 construction Methods 0.000 claims description 4
- 238000003062 neural network model Methods 0.000 abstract 1
- 230000006399 behavior Effects 0.000 description 13
- 238000000605 extraction Methods 0.000 description 4
- 230000003068 static effect Effects 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 210000003061 neural cell Anatomy 0.000 description 3
- 210000002569 neuron Anatomy 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 2
- 230000003542 behavioural effect Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 description 1
- 230000002159 abnormal effect Effects 0.000 description 1
- 210000004027 cell Anatomy 0.000 description 1
- 239000000306 component Substances 0.000 description 1
- 239000008358 core component Substances 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000003064 k means clustering Methods 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000003252 repetitive effect Effects 0.000 description 1
- 230000006403 short-term memory Effects 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Mathematical Physics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- Molecular Biology (AREA)
- Development Economics (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Game Theory and Decision Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (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
The invention discloses a method for classifying railway customers, which is characterized in that training is implemented to obtain a time sequence (Prophet) model and a long-short term memory network (LSTM) integrated model, freight ticket data of the railway customers in the freight process are respectively input into the Prophet model and the LSTM neural network model for processing, and respectively output results are integrated to obtain a customer delivery prediction time sequence; and extracting a characteristic space based on the historical delivery data and the predicted time sequence of the railway client, and clustering by adopting a nuclear power field to obtain a classification result of the railway client.
Description
Technical Field
The invention relates to the technical field of computers, in particular to a method for classifying railway customers.
Background
With the vigorous development of various transportation modes, the freight transportation market competition is increasingly violent, how to transport mass freight goods by using railways, deeply excavate the behavior characteristic information and the transportation demand of customers, analyze the loss tendency and the delivery potential of the customers, and is an important theoretical basis and technical support for realizing the online classification and the dynamic customer relationship management of the railway freight customers.
Currently, effective methods for classifying railroad customers mainly include: 1) a client value model and a comprehensive statistical index dividing method based on the life cycle of a railway freight client; 2) the method is a railway active customer clustering method based on customer static value index extraction of a customer value analysis model (RFM) or other improved models and clustering algorithms such as a k-means clustering algorithm (k-means). The method for classifying railway customers mainly aims at static historical data of customers for freight transportation, the loss of user characteristic extraction information is large, the relevance of delivery time series and individual customer trend of the railway customers in the freight transportation process are neglected, when the method faces massive and real-time railway freight transportation data and diversified customer demands, the problems of large calculated amount, low processing efficiency, poor adaptability and the like exist, and the method is difficult to meet the requirements of efficient mining analysis of the huge railway freight transportation customers.
The current method for classifying the behavioral event sequence information of the railway client mainly obtains classification results based on static historical data of freight transportation of the railway client, behavioral characteristic data of the railway client and set model parameters, and has the following defects: based on the clustering classification of static historical data of freight transportation performed by railway clients, data respectively presents sparsity and asymmetry in the high-dimensional time sequence of railway freight transportation user behaviors, so that the quality of a clustering classification result is difficult to judge; the clustering classification based on the behavior characteristic data of the railway client is to convert the original time sequence of the user behavior into a low-dimensional characteristic space vector for clustering classification, so that the dimensionality reduction of the sequence is realized, the interpretability is kept, and the adaptability is lower; the clustering classification based on the set model parameters is to convert the original time sequence of the railway customer behavior into a plurality of parameters of the model and perform clustering by using the parameters, but the clustering understandability of the user behavior parameters is relatively poor.
In the process of freight transportation of railway customers, the freight transportation behavior time sequence data has the characteristics of periodicity, difference, high volatility and strong seasonality, and the railway customers are clustered and classified by adopting any one of the three clustering methods, so that the railway customers are difficult to be classified accurately and dynamically in a comprehensive and accurate manner.
Disclosure of Invention
In view of this, the embodiments of the present invention provide a method for classifying railroad clients, which can accurately classify railroad clients comprehensively.
The embodiment of the invention is realized as follows:
a method of classifying railroad customers, comprising:
training to obtain a time sequence Prophet model and a long-short term memory network LSTM model;
respectively inputting the freight ticket data of the railway client in the freight process into the Prophet model and the LSTM model for processing, and outputting a result;
integrating the results respectively output from the Prophet model and the LSTM model to obtain a railway client delivery forecasting time sequence;
and after the obtained time sequence is constructed based on the characteristic space, clustering is carried out by adopting a nuclear power field to obtain a classification result of the railway client.
The Prophet model obtained by training carries out time sequence depth curve fitting based on railway client delivery on the freight data of the railway client;
and the LSTM model obtained by training excavates the dependency relationship of the time sequence of the railway client delivery in the ticket data of the railway client.
The integrating of the results respectively output from the Prophet model and the LSTM model comprises:
setting a first weight value corresponding to the Prophet model and a second weight value corresponding to the LSTM model, respectively, and adding a value obtained by multiplying the result output from the Prophet model by the first weight value and a value obtained by multiplying the result output from the LSTM model by the second weight value.
The construction of the obtained railway client delivery prediction time sequence based on the feature space comprises the following steps:
and processing the sample entropy SE, the statistical complexity SC and the complexity LZ on the obtained railway client delivery prediction time sequence.
The clustering by adopting the nuclear force field comprises the following steps: and calculating potential value entropy of the constructed railway client delivery prediction time sequence to obtain the optimal influence factor value of the minimum potential entropy so as to obtain the classification result of the railway client.
An apparatus for classifying railroad customers, comprising: a model setting unit, a processing unit, a feature space constructing unit and a clustering unit, wherein,
the model setting unit is used for training to obtain a Prophet model and an LSTM model;
the processing unit is used for respectively inputting the freight ticket data of the railway client in the freight process into the Prophet model and the LSTM model for processing, and combining the results respectively output from the Prophet model and the LSTM model to obtain a railway client delivery prediction time sequence;
the characteristic space construction unit is used for constructing the obtained railway client delivery prediction time sequence based on the characteristic space;
and the clustering unit is used for clustering the constructed railway customer delivery prediction time sequence by adopting a nuclear power field to obtain a classification result of the railway customer.
As seen from the above, the invention implements training to obtain a time sequence (Prophet) model and a long short term memory network (LSTM) model, the freight ticket data of the railway client in the freight process is respectively input into the Prophet model and the LSTM model for processing, and the results respectively output from the Prophet model and the LSTM model are merged to obtain the railway client delivery prediction time sequence; and constructing the obtained railway customer delivery prediction time sequence based on the characteristic space, and clustering by adopting a nuclear power field to obtain a classification result of the railway customer. Therefore, the railway examination can be comprehensively and accurately classified.
Drawings
FIG. 1 is a flow chart of a method for classifying railroad customers according to an embodiment of the present invention;
FIG. 2 is a schematic structural diagram of an LSTM model obtained by training according to an embodiment of the present invention;
fig. 3 is a schematic structural diagram of an apparatus for classifying railroad customers according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail below with reference to the accompanying drawings and examples.
In order to accurately classify railway customers comprehensively, the embodiment of the invention trains to obtain a Prophet model and an LSTM model, the freight ticket data of the railway customers in the freight process are respectively input into the Prophet model and the LSTM model for processing, and the results respectively output from the Prophet model and the LSTM model are combined to obtain a railway customer delivery prediction time sequence; and constructing the obtained railway customer delivery prediction time sequence based on the characteristic space, and clustering by adopting a nuclear power field to obtain a classification result of the railway customer.
The invention integrates a Prophet model and an LSTM model aiming at the behavior characteristic information of a railway client in the freight process, fully considers the railway freight characteristic and the freight characteristic of the railway client, extracts a railway client delivery forecasting time sequence by combining the trend, the periodicity, the holiday property and the front-back correlation of the behavior time sequence data of the railway client, constructs dynamic clustering based on a characteristic space and a nuclear power field, deeply analyzes the large-scale, high-latitude, multi-factor influence and unequal railway client delivery forecasting time sequence, fully excavates the internal implicit information of the behavior characteristic information of the railway client, realizes the dynamic characteristic extraction and accurate forecasting of the individual behavior of the railway client, and the dynamic online automatic classification of railway client groups based on real-time data is realized, and important data and technical support are provided for the accurate marketing and personalized service of railway clients.
Fig. 1 is a flowchart of a method for classifying railroad clients according to the present invention, which includes the following specific steps:
101, training to obtain a Prophet model and an LSTM model;
102, respectively inputting goods ticket data of a railway client in the freight process into the Prophet model and the LSTM model for processing, and outputting results;
103, integrating results respectively output from the Prophet model and the LSTM model to obtain a railway client delivery prediction time sequence;
and 104, constructing the obtained railway customer delivery prediction time sequence based on the feature space, and clustering by adopting a nuclear power field to obtain a classification result of the railway customer.
In the method, the freight ticket data is the behavior characteristic information of the railway client in the freight process.
Prior to step 102, the method further comprises:
and preprocessing the freight ticket data of the railway client in the freight process by adopting a deviation (min-max) standardization mode.
According to the method, the Prophet model obtained through training is used for fitting the freight ticket data of the railway client based on the railway client delivery time sequence data depth curve, and an output result is obtained. The Prophet model can have good adaptability to the festival effect and the trend change point in the goods ticket data of the railway client, has strong robustness aiming at missing values, trend transformation and a large number of abnormal values, and is formed by superposing a trend item, a period item and a festival item. The predicted value of the Prophet model at the time t is P (t), and is expressed as follows:
P(t)=g(t)+s(t)+h(t)+ε
the trend term g (t) is a core component of the model, and comprises different degrees of assumptions and smoothness adjustment parameters, and is used for fitting aperiodic changes in a time series and selecting change points from data to detect trend trends. The basic trend term g (t) is a logistic function:
where C represents the model capacity, k represents the growth rate, and g (t) tends towards C with the increase in time t.
s (t) is a period term which is approximated by a Fourier series to express a periodic component, the specific expression being as follows:
where T represents a certain fixed period and 2n represents the number of periods expected to be used in the model.
h (t) is a holiday term, and the influence of each holiday at different times is regarded as an independent model holiday term:
Z(t)=[1(t∈D1),...,1t(∈DL)
h(t)=Z(t)κ
wherein, κiRepresenting the influence of holidays in the window period on the predicted value; diRepresenting the ith virtual variable, if the time variable t belongs to the virtual variable, the virtual variable DiThe value is 1, otherwise 0; i denotes holidays, DiRepresents the time t contained in the window period; σ is an error term and follows a normal distribution, representing fluctuating residuals that the model does not predict.
The LSTM is a chain type of a repetitive neural network module, the model of the LSTM adopts a structure as shown in figure 2, and information is selectively transmitted through a sigmoid neural layer and a point-by-point multiplication operation through an input gate, a forgetting gate and an output gate.
Forgetting to gate means reading ht-1And xtOutputting a value between 0 and 1 to define each of the neural cell states Ct-1And further determining the lost information in the neuron cell state, wherein the formula is as follows:
ft=sigmoid(Wf·[ht-1,xt]+bf)
wherein h ist-1The output of the last neural cell unit, xtIndicating the input of the current cell.
The input gate is to determine how much new information to add to the current neural cell state, and the formula is:
it=sigmoid(Wi·[ht-1,xt]+bf)
ct=tanh(Wc·[ht-1,xt]+bc)
and the output gate is used for multiplying the output value of the running sigmoid layer and the output value of tanh based on the current neuron cell state and outputting the determined output part.
ot=sigmoid(Wo·[ht-1,xt]+bo)
ct=ft*ct-1+it*ct
ht=ot*tanh(ct)
Inputting the data characteristics at the time t into an input layer, outputting the results of LSTM structure nodes to a neuron of an output layer, calculating a back propagation error, continuously updating the network weight by using an Adam optimization algorithm to obtain a final hidden layer network, and outputting a prediction result L (t).
Defining the integrated Prophet-LSTM combined prediction model as follows:
Y(t)=ω1P(t)+ω2L(t)
ω1+ω2=1,t=1,2,…,N
wherein, t is the prediction time, and Y (t) is the result of adding the weight values of the Prophet model and the prediction data of the LSTM network. And calculating the optimal weight value formed by integrating the two models to obtain the Prophet-Lstm neural network client behavior prediction model.
In the method, the constructing the obtained railway client shipment forecasting time series based on the feature space comprises the following steps:
and carrying out global structure feature extraction of Sample Entropy (SE), Statistical Complexity (SC) and complexity (LZ) on the obtained railway client delivery forecasting time sequence for railway client delivery, constructing a clustered feature space, and then adopting dynamic clustering based on a nuclear power field on the feature space to obtain the optimal influence factor value of the minimum potential entropy so as to obtain the clustering result of the railway client.
Fig. 3 is a schematic structural diagram of an apparatus for classifying railroad clients according to an embodiment of the present invention, including: a model setting unit, a processing unit, a feature space constructing unit and a clustering unit, wherein,
the model setting unit is used for training to obtain a Prophet model and an LSTM model;
the processing unit is used for respectively inputting the freight ticket data of the railway client in the freight process into the Prophet model and the LSTM model for processing, and combining the results respectively output from the Prophet model and the LSTM model to obtain a railway client delivery prediction time sequence;
the characteristic space construction unit is used for constructing the obtained railway client delivery prediction time sequence based on the characteristic space;
and the clustering unit is used for clustering the constructed railway customer delivery prediction time sequence by adopting a nuclear power field to obtain a classification result of the railway customer.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like made within the spirit and principle of the present invention should be included in the scope of the present invention.
Claims (6)
1. A method of classifying railroad customers, comprising:
training to obtain a time sequence Prophet model and a long-short term memory network LSTM model;
respectively inputting the freight ticket data of the railway client in the freight process into the Prophet model and the LSTM model for processing, and outputting a result;
integrating the results respectively output from the Prophet model and the LSTM model to obtain a railway client delivery forecasting time sequence;
and after the obtained time sequence is constructed based on the characteristic space, clustering is carried out by adopting a nuclear power field to obtain a classification result of the railway client.
2. The method of claim 1, wherein the trained Prophet model performs a time series depth curve fit to the railway customer's invoice data based on the railway customer's shipment;
and the LSTM model obtained by training excavates the dependency relationship of the time sequence of the railway client delivery in the ticket data of the railway client.
3. The method of claim 1, wherein integrating the results output from the Prophet model and the LSTM model, respectively, comprises:
setting a first weight value corresponding to the Prophet model and a second weight value corresponding to the LSTM model, respectively, and adding a value obtained by multiplying the result output from the Prophet model by the first weight value and a value obtained by multiplying the result output from the LSTM model by the second weight value.
4. The method of claim 1, wherein constructing the resulting predicted time series of railway customer shipments based on feature space comprises:
and processing the sample entropy SE, the statistical complexity SC and the complexity LZ on the obtained railway client delivery prediction time sequence.
5. The method of claim 1, wherein the clustering using the nuclear force field is: and calculating potential value entropy of the constructed railway client delivery prediction time sequence to obtain the optimal influence factor value of the minimum potential entropy so as to obtain the classification result of the railway client.
6. An apparatus for classifying railroad customers, comprising: a model setting unit, a processing unit, a feature space constructing unit and a clustering unit, wherein,
the model setting unit is used for training to obtain a Prophet model and an LSTM model;
the processing unit is used for respectively inputting the freight ticket data of the railway client in the freight process into the Prophet model and the LSTM model for processing, and combining the results respectively output from the Prophet model and the LSTM model to obtain a railway client delivery prediction time sequence;
the characteristic space construction unit is used for constructing the obtained railway client delivery prediction time sequence based on the characteristic space;
and the clustering unit is used for clustering the constructed railway customer delivery prediction time sequence by adopting a nuclear power field to obtain a classification result of the railway customer.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911343327.8A CN111353523A (en) | 2019-12-24 | 2019-12-24 | Method for classifying railway customers |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911343327.8A CN111353523A (en) | 2019-12-24 | 2019-12-24 | Method for classifying railway customers |
Publications (1)
Publication Number | Publication Date |
---|---|
CN111353523A true CN111353523A (en) | 2020-06-30 |
Family
ID=71193985
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201911343327.8A Pending CN111353523A (en) | 2019-12-24 | 2019-12-24 | Method for classifying railway customers |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111353523A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113205368A (en) * | 2021-05-25 | 2021-08-03 | 合肥供水集团有限公司 | Industrial and commercial customer clustering method based on time sequence water consumption data |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104699767A (en) * | 2015-02-15 | 2015-06-10 | 首都经济贸易大学 | Large-scale ontology mapping method for Chinese languages |
CN104992244A (en) * | 2015-07-08 | 2015-10-21 | 中国民用航空总局第二研究所 | Airport freight traffic prediction analysis method based on SARIMA and RBF neural network integration combination model |
US20180174108A1 (en) * | 2016-12-19 | 2018-06-21 | Konolabs, Inc. | Method, system and non-transitory computer-readable recording medium for providing predictions on calendar |
CN109215346A (en) * | 2018-10-11 | 2019-01-15 | 平安科技(深圳)有限公司 | A kind of prediction technique, storage medium and the server of traffic transit time |
CN109992608A (en) * | 2019-03-26 | 2019-07-09 | 浙江大学 | A kind of multi-model fusion forecasting method and system based on frequency domain |
-
2019
- 2019-12-24 CN CN201911343327.8A patent/CN111353523A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104699767A (en) * | 2015-02-15 | 2015-06-10 | 首都经济贸易大学 | Large-scale ontology mapping method for Chinese languages |
CN104992244A (en) * | 2015-07-08 | 2015-10-21 | 中国民用航空总局第二研究所 | Airport freight traffic prediction analysis method based on SARIMA and RBF neural network integration combination model |
US20180174108A1 (en) * | 2016-12-19 | 2018-06-21 | Konolabs, Inc. | Method, system and non-transitory computer-readable recording medium for providing predictions on calendar |
CN109215346A (en) * | 2018-10-11 | 2019-01-15 | 平安科技(深圳)有限公司 | A kind of prediction technique, storage medium and the server of traffic transit time |
CN109992608A (en) * | 2019-03-26 | 2019-07-09 | 浙江大学 | A kind of multi-model fusion forecasting method and system based on frequency domain |
Non-Patent Citations (2)
Title |
---|
何晓旭: "时间序列数据挖掘若干关键问题研究" * |
葛娜 等: "Prophet-LSTM组合模型的销售量预测研究" * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113205368A (en) * | 2021-05-25 | 2021-08-03 | 合肥供水集团有限公司 | Industrial and commercial customer clustering method based on time sequence water consumption data |
CN113205368B (en) * | 2021-05-25 | 2022-11-29 | 合肥供水集团有限公司 | Industrial and commercial customer clustering method based on time sequence water consumption data |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN113962364B (en) | Multi-factor power load prediction method based on deep learning | |
Yoon et al. | Semi-supervised learning with deep generative models for asset failure prediction | |
CN112561156A (en) | Short-term power load prediction method based on user load mode classification | |
CN111324642A (en) | Model algorithm type selection and evaluation method for power grid big data analysis | |
Song et al. | An XGBoost algorithm for predicting purchasing behaviour on E-commerce platforms | |
CN113537469B (en) | Urban water demand prediction method based on LSTM network and Attention mechanism | |
CN110826237A (en) | Bayesian belief network-based wind power equipment reliability analysis method and device | |
CN112418476A (en) | Ultra-short-term power load prediction method | |
CN114066073A (en) | Power grid load prediction method | |
CN116187835A (en) | Data-driven-based method and system for estimating theoretical line loss interval of transformer area | |
CN116861924A (en) | Project risk early warning method and system based on artificial intelligence | |
González-Briones et al. | A review of the main machine learning methods for predicting residential energy consumption | |
El Maghraoui et al. | Smart energy management system: A comparative study of energy consumption prediction algorithms for a hotel building | |
CN117453764A (en) | Data mining analysis method | |
CN115409120A (en) | Data-driven-based auxiliary user electricity stealing behavior detection method | |
CN116128544A (en) | Active auditing method and system for electric power marketing abnormal business data | |
Wang et al. | Partition cost-sensitive CART based on customer value for Telecom customer churn prediction | |
CN115660725A (en) | Method for depicting multi-dimensional energy user portrait | |
CN113886454A (en) | Cloud resource prediction method based on LSTM-RBF | |
CN111353523A (en) | Method for classifying railway customers | |
CN117371573A (en) | Time sequence prediction method, device and medium based on TrAdaBoost-LSTM | |
Zhai et al. | Combining PSO-SVR and Random Forest Based Feature Selection for Day-ahead Peak Load Forecasting. | |
Lin | [Retracted] Big Data Technology in the Macrodecision‐Making Model of Regional Industrial Economic Information Applied Research | |
Heng et al. | A hybrid forecasting model based on empirical mode decomposition and the cuckoo search algorithm: a case study for power load | |
Wang et al. | A Novel Multi‐Input AlexNet Prediction Model for Oil and Gas Production |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200630 |