33 results about "ID3 algorithm" patented technology

The invention discloses a method and system for classifying data by adopting a decision tree. The method comprises the following steps of: parallel computing the information gain of each attribute in training data based on a MapReduce mechanism, and selecting optimum division decision attributes as nodes to construct the decision tree; based on the decision tree, classifying input data records. Aparallel decision tree ID3 algorithm is realized based on the MapReduce. A large data set can be processed, and the parallel efficiency is high. The parallel computing is realized for the nodes in the decision tree and the nodes in the same layer.

The invention provides a method for a power grid to detect the reasons of a voltage sag incident. The method includes the steps that historical data are extracted and subjected to discretization according to preset analysis parameters; the categorical attributes of voltage sag are set, and the historical data are classified to form a sample set, wherein the categorical attributes include the voltage sag phase, amplitude, duration time, occurring time and transmission characteristics; Apriori calculation is carried out on the sample set, a support degree value larger than a preset threshold value is obtained, Apriori calculation is carried out according to the support degree value, a credibility value lager than a preset second threshold value is obtained, and a voltage sag strong correlation rule knowledge base is formed; a decision-making tree is obtained through an ID3 algorithm; the reasons of current voltage sag of the power grid are determined according to the decision-making tree. The method and system for the power grid to detect the reasons of the voltage sag incident avoid depending on the waveform, excavate strong correlation rules in historical voltage sag incidents based on analysis of the historical data and through multiple characteristic values, and predict the possibility of voltage sag in the future.

The invention discloses a component classifying method based on net establishing software of a decision tree. A new component is added into a component bank of the net establishing software. A decision tree technology in data mining is used in classifying of the newly-added component. An ID3 algorithm based on information gain is used for carrying out analysis on the recommending degree of the newly-added component, the decision tree is established, and component classifying is completed. The decision tree technology is used in component classifying in the net establishing software, information gain is used for measuring component attribute values which are used as information amount provided for whole classifying, a classifying rule is visual, understanding and achieving are easy, and classifying efficiency is high.

The invention provides a method for predicting and evaluating concrete strength deterioration under an ocean environment. The method comprises the steps of detecting data of strength deterioration of concretes in different ratios under the ocean environment along with age through experiments; dividing the obtained operating data into a training group and a testing group, wherein factors which influence the concrete strength are used as a factor attribute set, and a strength deterioration state is used as a deterioration result attribute set; modeling a decision-making tree: selecting an output branch according to an attribute value of a corresponding attribute until a leaf node is reached, andoutputting an operating category for storing the leaf node as an analysis result; evaluating a model performance; and cutting branches of the established decision-making tree by use of a C4.5 algorithm. For complex concrete service environment deterioration conditions and final deterioration state responses, a final decision-making tree diagram is obtained by use of an ID3 algorithm, branches of the decision-making tree are reduced by use of the C4.5 optimization algorithm, the performance of the decision-making tree can be obviously improved, the concrete strength deterioration state under the ocean service condition is predicted and evaluated, an instant message of building strength deterioration can be obtained, and the risk caused by damage to building durability is assessed in advance.

The invention discloses a random forest-based airspace sector congestion degree prediction method and belongs to the field of air traffic congestion degree prediction. With the method adopted, the congestion degree of an airspace sector can be predicted scientifically and reasonably. The method of the invention includes following five steps of: reading historical data; preprocessing data; constructing a feature set; constructing a decision tree; and predicting the congestion level of the sector by using the random forest. According to the method, the five indicators, namely, sector capacity saturation, potential conflict number, sector aircraft density, sector aircraft average speed saturation and sector aircraft average distance, are processed; a fuzzy evaluation method is used to obtaincongestion levels corresponding to each time segment of the sector; an ID3 algorithm is used as a core algorithm to construct the decision tree; and samples are drawn and substituted into the decisiontree; classification is carried out layer by layer, and a prediction result is obtained; and three kinds of evaluation index data, such as prediction accuracy, prediction average absolute error, andprediction average percentage error, are calculated according to the prediction result; and the average value of each indicator is obtained, an whether prediction is accurate can be evaluated.

The invention provides a rock bust danger level prediction method based on a local weighting C4.5 algorithm and relates to the technical field of rock burst prediction. The method includes the steps of firstly, adopting an MDLP method for conducting discretization on continuous attribute data in sample data, then adopting a local weighting method for selecting a training set and calculating the weight of samples, utilizing the weight of the samples to calculate an information gain ratio of each attribute, and selecting sample attributes as root nodes of a C4.5 decision tree and splitting attributes of other branch nodes according to the information gain ratios; finally, adopting the weight of the samples to substitute the sample number to conduct pessimistic pruning on the created decisiontree, and correspondingly achieving prediction of rock burst dangers and the like in a predicted area. According to the provided rock bust danger level prediction method based on the local weightingC4.5 algorithm, the defect is overcome that the preference selection values have too many attributes when information gain is adopted for selecting node splitting attributes in an ID3 algorithm; an over-fitting problem is avoided, and the prediction accuracy of a model is high.

The invention relates to an emergent event classification and grading method, device and system based on decision trees and the Bayesian algorithm. The method comprises steps of S1, carrying out characteristic division on a preset grading and classification event library and constructing a training sample set; S2, according to the training sample set, using the ID3 algorithm, the C4.5 algorithm, and the CART algorithm to construct three decision tree classification and grading models; S3, according to the training sample set, constructing and training a Bayesian classifier; S4, carrying out key characteristic attribute extraction on events which are to be classified and graded; S5, according to event characteristic attributes, using the three decision tree models to carry out classification so as to obtain three classification results; and S6, according to the event characteristic attributes, using the Bayesian classifier to calculate probability of the categories of the three classification results in the S5, and taking the highest probability to be the final classification result. According to the invention, classification accuracy of a single algorithm can be improved; and a disadvantage of difficulty in predicting continuous fields in a decision tree algorithm is effectively overcome.

The invention discloses a power business collaborative classification method and system based on an ID3 decision tree algorithm. The power business collaborative classification method comprises the following steps: obtaining a power business collaborative related database, and extracting a sample set S from the power business collaborative related database; extracting an index set A, wherein the index set A contains indexes for evaluating the business collaboration data; calculating the information entropy and the information gain of each index for the sample set S based on an ID3 algorithm soas to select a proper root node and a proper intermediate node; and constructing a decision tree according to the selected root node; evaluating and selecting each service cooperation scheme based onthe decision tree. The invention further discloses a corresponding system. According to the power business collaborative classification method, information entropy and information gain calculation isadopted, and the calculation amount is relatively small, and the classification accuracy is high, and the power business collaborative classification method is applied to cooperative data calculationand analysis of services such as power outsourcing, and optimal division characteristics are selected as nodes to generate a decision-making tree and perform data classification, and classification is rapid and good in effect, and cooperative management of the services such as power outsourcing is effectively achieved.

The invention discloses a sea-surface oil spill monitoring threshold setting method based on an ID3 algorithm and a neural network and mainly solves the problem that the prior art cannot well apply tocomplex monitoring environments and changing detection distance. The method includes: using a monitoring system to acquire the environment information such as weather, tide, sun altitude and ultraviolet of a to-be-monitored sea area through the Internet; selecting and using an ID3 decision tree monitoring threshold or a neural network monitoring threshold according to a set time limit; performingmatched filtering on the monitoring data of the to-be-monitored sea surface to acquire a data maximum value, judging whether the maximum value exceeds the monitoring threshold or not, if so, transmitting an alarm, and if not, judging that oil spill does not occur on the sea surface; after staff receive the alarm, by the staff, judging whether oil spill actually occurs or not through a real-time image, if so, processing the oil spill immediately, and if not, manually modifying the threshold, and reconstructing an ID3 decision tree or training the neural network. By the method applicable to sea-surface oil spill monitoring, monitoring precision is increased.

The embodiment of the invention provides a park evaluation method and device, a readable medium and electronic equipment, and the method comprises the steps: receiving a query request which carries park information; based on the park information, calling the corresponding description data from the evaluation database, and performing calculation through an ID3 algorithm to obtain evaluation data ofthe corresponding park; and based on the evaluation data, performing prediction analysis through the trained caffe model to obtain an evaluation report of the corresponding park. According to the technical scheme, the sample capacity is large, the model can be updated in the evaluation process, the park evaluation efficiency, objectivity and accuracy are improved on the whole, the workload is reduced, the problems existing in current park construction and development can be found, and decision support is provided for park construction.

The invention discloses a data classification method based on an ID3 algorithm. The method comprises the following steps: optimizing an information gain Gain (Ai) obtained by an ID3 algorithm by utilizing an equalization coefficient R (Ai) to obtain an optimized information gain Gain (Ai) new, obtaining a root node and branch nodes of a decision tree according to the optimized information gain Gain (Ai) new, and classifying attributes Ai. By introducing the equalization coefficient R (Ai) and an attribute deviation threshold T, the measurement and control of the multi-value deviation degree are realized, the problem of attributes with more information gain deviation values can be avoided, and the prediction accuracy and the practicability and effectiveness of the ID3 algorithm are furtherimproved.

The invention discloses a storage optimization method for a Hadoop distributed file system, and the method specifically comprises the following steps: firstly, selecting an INFO-level log file which comprises a specific execution timestamp and file name information, and obtaining an access record and a deletion record of the INFO-level log file; extracting and sorting all information containing keywords in the IFNO level log, and then sorting and numbering according to timestamps; then determining feature tags, selecting features, constructing feature vectors, and forming a sample set of a training file elimination model; selecting three feature values of the feature vector as three classification nodes of a decision tree in sequence, establishing the decision tree by adopting an ID3 algorithm, and constructing a file elimination model by the decision tree; and finally, predicting the reusability of the file by using the established file elimination model. According to the method, the storage efficiency of the distributed file system is optimized, the data storage scale is reduced, and the storage efficiency of the HDFS is improved.

The invention provides an intelligence data fusion method and device, a Smart rule is trained and generated by adopting a decision tree ID3 algorithm, and fusion of network intelligence data is realized by performing entity extraction, entity classification, attribute recognition and attribute extraction on original network intelligence data and automatically selecting a fusion rule. The invention mainly aims to solve the problems of low efficiency and uneven fusion effect of the existing information data fusion, realize the efficient, rapid and standardized fusion of the network information data, and reduce the dependency of the network information data fusion on the knowledge of field experts.

The invention discloses an RLID3 data classification method based on a decision tree optimization rate, and relates to an RLID3 data classification method based on a decision tree optimization rate, for the purpose of solving the problem of low classification accuracy of a conventional decision tree. The concrete process comprises the following steps: step one, optimizing information gain to obtain the decision tree optimization rate; and step two, based on the decision tree optimization rate, obtaining an RLID3 algorithm, wherein the RLID3 algorithm involves converting the information gain into the decision tree optimization rate in the process of an ID3 algorithm. The method is applied to the field of classification.

The invention discloses an ID3 algorithm-based black and odorous water body remote sensing image identification method. The method comprises the steps of 1, high-resolution remote sensing image data collection and processing, 2, water body extraction and chemical index measurement, 3, spectral feature selection, 4, construction of the ID3 algorithm and a classification decision tree, and 5, precision verification of a verification set. The objective and standard threshold value calculation method is determined based on ID3, the requirement for black and odorous water body information extraction in the image is met, good universality is achieved in practical application, meanwhile, multi-feature classification is added in the analysis process, classification is further expanded through multiple features, the result is made to better conform to objective features of a research area. The black and odorous water body extraction precision can be effectively improved, and the robustness and universality of black and odorous water body extraction of the high-resolution remote sensing image are improved.

The invention relates to a decision tree generation method based on an ID3 algorithm, which improves the ID3 algorithm. Means + + algorithm discretizes the values of the continuous attributes in the dataset, then the importance SGA (a, P, A) of each conditional attribute is calculated, the attributes with great importance are selected as splitting points, iterating is performed repeatedly until all conditional attributes are used as splitting nodes, and finally pruning is performed to form a decision tree.

The invention provides a static testing null pointer reference defect false-positive recognition method. According to the software static testing null pointer reference defect false-positive problem, a static testing defect report of to-be-tested software and null pointer reference defect knowledge in a software historical warehouse are extracted; null pointer reference defect initiating conditions are extracted in a null pointer reference defect mode and compared with the null pointer reference defect knowledge, so that a null pointer reference defect associate attribute group is determined, and null pointer reference defect data sets are constructed; the null pointer reference defect data sets are classified through the ID3 algorithm based on the rough set theory attribute importance, classification results are used for null pointer reference defect false-positive recognition, and true null pointer reference defects are confirmed. The method is combined with the null pointer reference defect knowledge and the ID3 algorithm based on the rough set theory attribute importance for static testing null pointer reference defect false-positive recognition, the detection efficiency and stability of the static testing null pointer reference defects are improved, and the null pointer reference defect confirmation cost is reduced.

The invention discloses a power system fault classification method and system based on an ID3 algorithm, and belongs to the technical field of data analysis. The method comprises the steps of firstlyobtaining a data set, then processing the data set by utilizing an ID3 algorithm to obtain a decision tree, and then classifying data of the data set according to the decision tree to obtain a classification result. The system comprises a tree building module and a classification module, the tree building module is connected with the classification module, and the tree building module is used forbuilding a decision tree by utilizing an ID3 algorithm according to a data set. The invention overcomes the defects of low accuracy and low diagnosis efficiency of power system fault diagnosis in theprior art, provides the ID3 algorithm-based power system fault classification method and system, and can quickly diagnose and classify the fault data through the decision tree, thereby improving the diagnosis efficiency of the power system fault and further improving the accuracy of power system fault diagnosis.

The invention discloses a man-machine relationship verification method and device based on equipment use conditions. The method comprises the following steps: obtaining an equipment use condition database, and extracting a training set D from the equipment use condition database; extracting a feature set A of the training set D, wherein the feature set A contains features for judging the use condition of the equipment; calculating the empirical conditional entropy and the information gain of each feature in the feature set A to the training set D based on an ID3 algorithm so as to select a proper root node and a proper intermediate node; constructing a decision tree according to the selected root node and the intermediate node; and analyzing whether the equipment is frequently used or notand whether the equipment is replaced and used arbitrarily or not based on the decision tree so as to adjust and maintain the requirements of the equipment. The invention further discloses a corresponding device. Whether the equipment is frequently used or not and whether the behavior of replacing the equipment without permission occurs or not can be discovered in time through equipment feature identification calculation so that the information security and accuracy are improved, and the problems that the equipment is idle and the user of the equipment can be replaced without permission are solved.