A high-dimensional feature selection method for an information gain hybrid neighborhood rough set
A feature selection method and a neighborhood rough set technology, applied in the field of image processing, can solve problems such as affecting the reduction effect and losing important information, and achieve the effect of reducing time complexity, ensuring scientificity, and improving accuracy.
Active Publication Date: 2019-06-25
NINGXIA MEDICAL UNIV
View PDF16 Cites 7 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
However, Pawlak RS can only deal with nominal variables, and the data in practical applications are often continuous numerical variables. Although the discretized data set can be adapted to the construction of the equivalent class of the RS algorithm, it may also lose important information and different discrete The reduction strategy also affects the reduction effect
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment
[0044] (1) Data acquisition
[0045] The data comes from the General Hospital of Ningxia Medical University. The data of each case includes clinical diagnosis results, imaging data, examination findings, etc. The clinical conclusion is the standard for diagnosing benign and malignant lung tumors. In order to avoid insufficient model training due to insufficient data, this study is not limited to a certain type of lung tumor. Therefore, 3000 cases of lung tumor data were obtained, including CT data of 1500 cases of malignant lung tumors and 1500 cases of benign lung tumors.
[0046] (2) Data preprocessing
[0047] The CT images of benign and malignant lung tumors were obtained from the DICOM files according to the inspection conclusions in the medical order of each case, and the images were numbered in sequence, and the false colors were removed and converted into grayscale images. In the grayscale image, centering on the lesion marked by the radiologist, the sub-image with s...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention discloses a high-dimensional feature selection method for an information gain hybrid neighborhood rough set. The method comprises the following specific steps: step 1, data preprocessing; Step 2, image segmentation; Step 3, feature extraction; 4, feature normalization; 5, feature selection based on information gain; 6, feature selection based on the domain rough set; And step 7, performing of classification and identification on the two reduction results. The invention provides a high-dimensional feature selection method for an information gain hybrid neighborhood rough set, andthe feasibility of a two-stage reduction algorithm is analyzed from the theory level. The algorithm can improve the accuracy of the algorithm, effectively reduce the time complexity, comprehensively compare the performance of the high-dimensional feature selection algorithm constructed by different methods, ensure the superiority of the method, ensure the scientificity of the result from the step-by-step selection of the model method, and have a certain reference value for the identification of benign and malignant lung tumors.
Description
technical field [0001] The invention relates to the technical field of image processing, and more specifically relates to a high-dimensional feature selection method of mixed neighborhood rough sets with information gain. Background technique [0002] Information gain (information gain, IG) and rough set (rough set, RS) are two commonly used algorithms for feature selection. IG is an index to measure how much information is provided for the classifier when a certain feature is included or not included. Find out the amount of information provided by each feature to the classifier, then sort from large to small, and select the top K features according to certain rules, so as to achieve the purpose of feature selection using information gain. The calculation complexity of IG for feature selection is low, and only a single operation is required, so the operation efficiency is high, and redundant, irrelevant and noise features can be effectively eliminated. However, there are st...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/62G06K9/34
Inventor 陆惠玲周涛张飞飞梁蒙蒙杨健
Owner NINGXIA MEDICAL UNIV