CN111178441A - Lithology identification method based on principal component analysis and full-connection neural network - Google Patents
Lithology identification method based on principal component analysis and full-connection neural network Download PDFInfo
- Publication number
- CN111178441A CN111178441A CN201911413844.8A CN201911413844A CN111178441A CN 111178441 A CN111178441 A CN 111178441A CN 201911413844 A CN201911413844 A CN 201911413844A CN 111178441 A CN111178441 A CN 111178441A
- Authority
- CN
- China
- Prior art keywords
- data
- principal component
- fcnn
- neural network
- component analysis
- 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
Images
Classifications
-
- 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
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2135—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on approximation criteria, e.g. principal component analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
-
- 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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Health & Medical Sciences (AREA)
- Probability & Statistics with Applications (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a lithology identification method based on principal component analysis and a fully-connected neural network, which comprises the following steps of: inputting data to be tested into a PCA-FCNN model, and obtaining a lithology recognition result after processing through the PCA-FCNN model; the establishment of the PCA-FCNN model comprises the following steps: s1, selecting and normalizing a training sample; taking logging curves with different lithological characteristics as input data, selecting data from the input data as training samples, and performing normalization processing after disordering the data in the training samples; s2, performing feature dimensionality reduction on the normalized data; performing characteristic dimensionality reduction on the normalized data through a principal component analysis method, and selecting n principal components according to a dimensionality reduction result; s3, establishing a neural network by adopting an FCNN algorithm; and establishing n layers of FCNN neural networks according to the number of the main components, wherein the activation function of the front n-1 layer is set as a Relu function, and the nth layer adopts a softmax function. The method can obviously improve the accuracy of lithology identification.
Description
Technical Field
The invention relates to the technical field of lithology identification, in particular to a lithology identification method based on principal component analysis and a fully-connected neural network.
Background
Well log lithology recognition is the first job of interpretation of well logs, the importance of which is self evident. The conventional lithology interpretation methods mainly comprise methods such as cross-plot analysis, multivariate statistical chart analysis, formation element logging and the like, and the methods are widely applied to lithology identification at home and abroad. The cross-plot analysis method is simpler, but needs a plurality of well logging curves which can reflect lithology; the multivariate statistical analysis method has small workload and high speed, but various parameters need to be adjusted in the lithology identification process, and the error is large; formation element logging can identify lithology of rock compositions from a geochemical perspective, but this method is too expensive and consumes a great deal of cost. Therefore, there is a need for a lithology identification method that can not only consume less manpower and material resources, but also obtain an accurate lithology identification result.
The development of artificial intelligence, particularly deep learning, brings a new approach to lithology identification. The Full Connection Neural Network (FCNN) is an important algorithm for deep learning and has wide application in many scientific fields, and as FCNN belongs to one of supervised learning, the selection of proper training set features and the accuracy of training data and label data have an important influence on the result of the neural network.
Therefore, it is an urgent problem to those skilled in the art to develop a lithology recognition method based on principal component analysis and fully-connected neural network.
Disclosure of Invention
In view of the above, the invention provides a lithology identification method based on principal component analysis and a fully-connected neural network, which is used for solving the problems of poor identification effect and low accuracy in identification of complex lithology.
In order to achieve the purpose, the invention adopts the following technical scheme:
a lithology recognition method based on principal component analysis and a fully-connected neural network comprises the following steps:
inputting data to be tested into a PCA-FCNN model, and obtaining a lithology recognition result after processing through the PCA-FCNN model;
the establishment of the PCA-FCNN model comprises the following steps:
s1, selecting and normalizing a training sample; taking logging curves with different lithological characteristics as input data, selecting data from the input data as training samples, and performing normalization processing after disordering the data in the training samples;
s2, performing feature dimensionality reduction on the normalized data; performing characteristic dimensionality reduction on the normalized data through a principal component analysis method, and selecting n principal components according to a dimensionality reduction result;
s3, establishing a neural network by adopting an FCNN algorithm; and establishing n layers of FCNN neural networks according to the number of the main components, wherein the activation function of the front n-1 layer is set as a Relu function, and the nth layer adopts a softmax function.
Preferably, the normalization process specifically includes the following:
respectively carrying out normalization processing on data in the training samples according to lithological characteristics;
wherein the lithological characteristics include: acoustic moveout, density, resistivity, natural gamma and natural potential.
Preferably, S2 specifically includes the following:
and performing feature dimensionality reduction on the normalized data through a principal component analysis method, selecting the dimensionality number of the output data based on the sequence arrangement of the proportion of each principal component variance in the total variance after dimensionality reduction, and removing the principal component with a small variance proportion.
Preferably, a Dropout algorithm is added to the front (n-1) layer of the FCNN neural network.
Through the technical scheme, compared with the prior art, the invention discloses a lithology recognition method based on principal component analysis and a fully-connected neural network, firstly, the lithology recognition is realized by combining the principal component analysis with the fully-connected neural network, wherein the principal component analysis can effectively eliminate the correlation degree between data, and the accuracy can be obviously improved by combining the data with the fully-connected neural network, secondly, the data is normalized in the invention, the possibility of error generation of a classification result caused by the difference of the number of samples is effectively reduced, the accuracy of the method is further improved, in addition, a Relu function is adopted in the FCNN of the invention, when the function is input as a positive number (for most of input z spaces), the gradient disappearance problem does not exist, and the calculation speed is high, the working efficiency of the invention is effectively improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the provided drawings without creative efforts.
FIG. 1 is a cross-sectional view of a well log provided by an embodiment of the present invention;
FIG. 2 is a flow chart of the PCA-FCNN model-based lithology recognition provided by the present invention;
FIG. 3 is a histogram of covariance ratio bars of principal components after dimensionality reduction, according to an embodiment of the present invention;
FIG. 4 is a graph of training error curves and accuracy provided by an embodiment of the present invention;
fig. 5 is a comparison graph of the actual measurement result and the lithology identification result provided by the embodiment of the invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The embodiment of the invention discloses a lithology identification method based on principal component analysis and a fully-connected neural network, which comprises the following steps:
inputting data to be tested into a PCA-FCNN model, and obtaining a lithology recognition result after processing through the PCA-FCNN model;
the establishment of the PCA-FCNN model comprises the following steps:
s1, selecting and normalizing a training sample; taking logging curves with different lithological characteristics as input data, selecting data from the input data as training samples, and performing normalization processing after disordering the data in the training samples;
s2, performing feature dimensionality reduction on the normalized data; performing characteristic dimensionality reduction on the normalized data through a principal component analysis method, and selecting n principal components according to a dimensionality reduction result;
s3, establishing a neural network by adopting an FCNN algorithm; and establishing n layers of FCNN neural networks according to the number of the main components, wherein the activation function of the front n-1 layer is set as a Relu function, and the nth layer adopts a softmax function.
In order to further implement the above technical solution, the normalization process specifically includes the following steps:
respectively carrying out normalization processing on data in the training samples according to lithological characteristics;
wherein the lithological characteristics include: acoustic moveout, density, resistivity, natural gamma and natural potential.
In order to further implement the above technical solution, S2 specifically includes the following contents:
and performing feature dimensionality reduction on the normalized data through a principal component analysis method, selecting the dimensionality number of the output data based on the sequence arrangement of the proportion of each principal component variance in the total variance after dimensionality reduction, and removing the principal component with a small variance proportion.
It should be noted that:
the selection condition of selecting the dimension number is that if the variance ratio of the first k features exceeds 95%, the k features are selected as the main components of the output.
In order to further implement the technical scheme, a Dropout algorithm is added into the front (n-1) layer of the FCNN neural network.
It should be noted that:
the Dropout algorithm is added to some or all of the previous (n-1) layers followed by the last layer, which is the softmax function, without the Dropout algorithm.
The first embodiment is as follows:
taking a new landscape mine field located in the west of the Yangquan mining area of the Qin basin as an example:
lithology is mainly divided into four categories: sandstone (siltstone, sandstone, and siltstone), mudstone, sandy mudstone, limestone, and coal. The well logging curves of the research area mainly comprise: acoustic time difference (AC), Density (DEN), Resistivity (Resistivity), natural Gamma (GR), and natural potential (SP). The cross-plot of the log is shown in FIG. 1.
It can be known from the figure that, the sandy mudstone contains the components of sandstone and mudstone and is judged as the content percentage of sandstone and mudstone, and the sandstone includes fine sandstone, medium sandstone, coarse sandstone and siltstone, the logging response of the coarse sandstone and the siltstone is similar to that of the sandy mudstone, so the sandstone is displayed on the logging display, the mudstone and the sandy mudstone are connected together, and no good distinguishing effect occurs, because the logging curve of the coal seam is characterized by high Resistivity, high acoustic wave time difference and low density natural gamma value, but the difference of the value ranges of the logging response values of four lithologies in the logging response is not obvious in the intersection graph, so the acoustic wave time difference (AC), the Density (DEN), the Resistivity (Resistivity), the natural Gamma (GR) and the natural potential (SP) curve have no good identifying effect on the lithology, therefore, the method utilizes a fully-connected deep neural network (PCA-FCNN) optimized by principal component analysis to perform labeled supervised learning on five curves of sound wave time difference (AC), Density (DEN), Resistivity (Resistivity), natural Gamma (GR) and natural potential (SP), divides lithology into 5 types (coal, sandstone, limestone, sandy mudstone and mudstone) more finely, and deeply excavates the potential of logging data, and a specific flow chart is shown in FIG. 2.
The sample data selects logging data and lithology interpretation data of 550-670 m logging in a new prospect 3-167, and because the difference of the sample number of each lithology is large, the error of a classification result caused by the difference of the sample number is not prevented, the same amount of sample number is selected from each lithology for training, and the specific conditions are as shown in the table below.
TABLE 1 sample data screening
As can be seen from the data in the table, the logging data of coal and limestone is too rare compared with other lithologies, so that 200 groups of data are randomly extracted from the sample data of each lithology by taking the logging data point of the limestone as a reference to form a data set.
In order to enable the loss value of the data in the training set to be rapidly reduced and prevent the training result from being influenced by the overlarge range of some characteristic value domains in the training set, the log data in the data set are subjected to disordering sequence processing and normalization processing according to a formula I.
Wherein max is the maximum value of the characteristic sample data, and min is the minimum value of the characteristic of the sample data.
The results after normalization are shown in table 2.
TABLE 2 normalized post log data
The normalized data set was cut with 70% of the data as the training set and 30% as the test set.
The normalized data is subjected to principal component analysis, and the number of dimensions of the output data is selected based on the order arrangement of the magnitude of the total variance ratio of the dimensionality-reduced variance, and if the variance ratio of the principal component is larger, the higher the importance degree of the principal component is, as shown in fig. 3.
As can be seen from the above figure, the variance ratios of the principal components after dimensionality reduction are 0.59265653, 0.197077, 0.1017802, 0.06600137 and 0.04248491, and since the sum of the first four principal components has reached 95.7%, the first four principal components (PCA1, PCA2, PCA3 and PCA4) are selected as the input features of the FCNN in this study, and the data are shown in table 3.
TABLE 3 principal Components analysis results
To obtain a well-behaved lithology recognition model and based on 4 principal components (PCA1, PCA2, PCA3, PCA4) obtained by principal component analysis, we chose to use the FCNN algorithm to build a four-layer neural network (as in table 4), and the neurons of each layer are set to 10, 20, 30, 5, respectively. In order to obtain better training effect, the activation function of the first three layers is selected to be set as the Relu function. Because the main lithology of the work area well logging is divided into 5 types, the number of the neurons in the last layer is 5, and the softmax function commonly used for the classification problem is selected as the activation function. On the basis, a Dropout algorithm is added behind the second layer and the third layer of the neural network, namely, neurons in a certain proportion are shielded in forward propagation, so that overfitting of the model is prevented, the accuracy of the model is improved, and 10% of the neurons which are randomly shielded are selected in the model.
TABLE 4 neural network architecture
The learning rate of the model is set to 0.01, the number of times of training is set to 500, and the training result is shown in fig. 4.
After 500 times of training, the final loss value is reduced to 0.2213, and the accuracy is improved to 93.45%. The test set is input into the trained model with an accuracy of 91%.
And 3-167 well logging 520 m-550 m lithology histogram data is selected to verify the PCA-FCNN lithology recognition model, the result is shown in figure 5, and the result shows that the PCA-FCNN lithology recognition model has a good effect and can accurately recognize the lithology.
To test the advantages of the PCA-FCNN algorithm, the FCNN algorithm was chosen for comparison, and five curves of acoustic time difference (AC), Density (DEN), Resistivity (Resistivity), natural Gamma (GR) and natural potential (SP) were chosen as input data. Constructing a neural network with 4 layers, 10 neurons, 20 neurons, 30 neurons and 8 neurons, Relu and Softmax, and adopting 10% Dropout operation on the second layer and the third layer. The learning rate was set to 0.01, after 500 trainings, the loss value decreased to 0.2935, the accuracy increased to 88.18%, the accuracy of the test set was 83%, and the comparison is shown in table 5. The results show that the accuracy of the PCA-FCNN algorithm is higher than that of the FCNN algorithm.
TABLE 5 model comparison
The embodiments in the present description are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other. The device disclosed by the embodiment corresponds to the method disclosed by the embodiment, so that the description is simple, and the relevant points can be referred to the method part for description.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (4)
1. A lithology recognition method based on principal component analysis and a fully-connected neural network is characterized by comprising the following steps:
inputting data to be tested into a PCA-FCNN model, and obtaining a lithology recognition result after processing through the PCA-FCNN model;
the establishment of the PCA-FCNN model comprises the following steps:
s1, selecting and normalizing a training sample; taking logging curves with different lithological characteristics as input data, selecting data from the input data as training samples, and performing normalization processing after disordering the data in the training samples;
s2, performing feature dimensionality reduction on the normalized data; performing characteristic dimensionality reduction on the normalized data through a principal component analysis method, and selecting n principal components according to a dimensionality reduction result;
s3, establishing a neural network by adopting an FCNN algorithm; and establishing n layers of FCNN neural networks according to the number of the main components, wherein the activation function of the front n-1 layer is set as a Relu function, and the nth layer adopts a softmax function.
2. The lithology identification method based on principal component analysis and fully-connected neural network as claimed in claim 1, wherein the normalization process specifically comprises the following steps:
respectively carrying out normalization processing on data in the training samples according to lithological characteristics;
wherein the lithological characteristics include: acoustic moveout, density, resistivity, natural gamma and natural potential.
3. The lithology identification method based on principal component analysis and fully-connected neural network as claimed in claim 1, wherein S2 specifically includes the following contents:
and performing feature dimensionality reduction on the normalized data through a principal component analysis method, selecting the dimensionality number of the output data based on the sequence arrangement of the proportion of each principal component variance in the total variance after dimensionality reduction, and removing the principal component with a small variance proportion.
4. The method for lithology recognition based on principal component analysis and fully-connected neural network as claimed in claim 1, wherein Dropout algorithm is added in the front (n-1) layer of the FCNN neural network.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911413844.8A CN111178441A (en) | 2019-12-31 | 2019-12-31 | Lithology identification method based on principal component analysis and full-connection neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911413844.8A CN111178441A (en) | 2019-12-31 | 2019-12-31 | Lithology identification method based on principal component analysis and full-connection neural network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111178441A true CN111178441A (en) | 2020-05-19 |
Family
ID=70652367
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911413844.8A Pending CN111178441A (en) | 2019-12-31 | 2019-12-31 | Lithology identification method based on principal component analysis and full-connection neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111178441A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111783825A (en) * | 2020-05-26 | 2020-10-16 | 中国石油天然气集团有限公司 | Well logging lithology identification method based on convolutional neural network learning |
| CN111899338A (en) * | 2020-08-05 | 2020-11-06 | 芯元(浙江)科技有限公司 | Method, device and system for three-dimensional modeling of stratum lithology of coverage area |
| CN112686259A (en) * | 2020-12-16 | 2021-04-20 | 中国石油大学(北京) | Rock image intelligent identification method and device based on deep learning and storage medium |
| CN112990320A (en) * | 2021-03-19 | 2021-06-18 | 中国矿业大学(北京) | Lithology classification method and device, electronic equipment and storage medium |
| CN113343574A (en) * | 2021-06-21 | 2021-09-03 | 成都理工大学 | Mishrif group lithology logging identification method based on neural network |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2019149376A1 (en) * | 2018-02-02 | 2019-08-08 | Toyota Motor Europe | Method and system for processing input data using a neural network and normalizations |
| CN110618082A (en) * | 2019-10-29 | 2019-12-27 | 中国石油大学(北京) | Reservoir micro-pore structure evaluation method and device based on neural network |
-
2019
- 2019-12-31 CN CN201911413844.8A patent/CN111178441A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2019149376A1 (en) * | 2018-02-02 | 2019-08-08 | Toyota Motor Europe | Method and system for processing input data using a neural network and normalizations |
| CN110618082A (en) * | 2019-10-29 | 2019-12-27 | 中国石油大学(北京) | Reservoir micro-pore structure evaluation method and device based on neural network |
Non-Patent Citations (1)
| Title |
|---|
| 张国英等: "一种基于主成分分析的BP神经网络在岩性识别中的应用", 《北京石油化工学院学报》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111783825A (en) * | 2020-05-26 | 2020-10-16 | 中国石油天然气集团有限公司 | Well logging lithology identification method based on convolutional neural network learning |
| CN111899338A (en) * | 2020-08-05 | 2020-11-06 | 芯元(浙江)科技有限公司 | Method, device and system for three-dimensional modeling of stratum lithology of coverage area |
| CN112686259A (en) * | 2020-12-16 | 2021-04-20 | 中国石油大学(北京) | Rock image intelligent identification method and device based on deep learning and storage medium |
| CN112686259B (en) * | 2020-12-16 | 2023-09-26 | 中国石油大学(北京) | Rock image intelligent recognition method and device based on deep learning and storage medium |
| CN112990320A (en) * | 2021-03-19 | 2021-06-18 | 中国矿业大学(北京) | Lithology classification method and device, electronic equipment and storage medium |
| CN113343574A (en) * | 2021-06-21 | 2021-09-03 | 成都理工大学 | Mishrif group lithology logging identification method based on neural network |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111178441A (en) | Lithology identification method based on principal component analysis and full-connection neural network | |
| CN110346831B (en) | Intelligent seismic fluid identification method based on random forest algorithm | |
| Silva et al. | Petrofacies classification using machine learning algorithms | |
| CN108952699B (en) | Intelligent stratum lithology identification method in complex geological drilling process | |
| Yu et al. | Synthetic sonic log generation with machine learning: A contest summary from five methods | |
| CN105572747B (en) | A method of identification waterflooding reservoir clastic rock lithology in the areas Fu Jia with high salt | |
| CN111596354A (en) | Seismic reservoir prediction method considering space constraint under machine learning framework | |
| CN114492521A (en) | Intelligent lithology while drilling identification method and system based on acoustic vibration signals | |
| Chang et al. | SegLog: Geophysical logging segmentation network for lithofacies identification | |
| Geng et al. | Channel attention-based static-dynamic graph convolutional network for lithology identification with scarce labels | |
| CN104570109A (en) | Method for reservoir petroleum gas prediction | |
| Wang et al. | Identification protocols for horizontal‐to‐vertical spectral ratio peaks | |
| CN113033637B (en) | Integrated learning lithofacies prediction method based on rock physical sample enhancement | |
| Li et al. | Lithology classification based on set-valued identification method | |
| Samanta et al. | Sparse data division using data segmentation and Kohonen network for neural network and geostatistical ore grade modeling in Nome offshore placer deposit | |
| Kakouei et al. | Lithological facies identification in Iranian largest gas field: A comparative study of neural network methods | |
| Chikhi et al. | Hybrid neural network methods for lithology identification in the Algerian Sahara | |
| CN112147679B (en) | Lithology prediction method and device based on elastic parameters under fuzzy logic framework | |
| CN115809411A (en) | Improved decision tree lithology identification method based on logging data | |
| Wang et al. | Borehole lithology modelling with scarce labels by deep transductive learning | |
| de Oliveira et al. | Ensemble of heterogeneous classifiers applied to lithofacies classification using logs from different wells | |
| Wang et al. | Auto recognition of carbonate microfacies based on an improved back propagation neural network | |
| CN112257789A (en) | Method for identifying surrounding rock grade | |
| Fung et al. | Lithology classification using self-organising map | |
| Emery et al. | The Pitfalls and Insights of Log Facies Classification using Machine Learning |
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 |