Dissertation > Excellent graduate degree dissertation topics show

Research Based on Multi_Map Graphics Primitives and Classification of Multi_Dimensional

Author: LiuYanJu
Tutor: HongWenXue
School: Yanshan University
Course: Biomedical Engineering
Keywords: Data Visualization Radial coordinate Figure Feature Fusion Hierarchical Feature Selection
CLC: TP391.41
Type: Master's thesis
Year: 2010
Downloads: 106
Quote: 0
Read: Download Dissertation

Abstract


The multi-dimensional data classification is an important research topic in the pattern recognition. Today, the classification algorithm there are some problems: the traditional classification algorithm requires a lot of computing, the more complex problems of classification and identification of the target, the interpretability of the classification results, the classification process agnostic and other issues. In order to solve the above problems, this paper studies how the characteristics of multi-map graphics primitive said and characteristics of fusion the characteristics extraction technology as a means to reduce the calculation of the consideration for the classification algorithm and classification results visualization technology, the classification process visualization, based on multi-figure graphics and proposed The general characteristics of primitive multi-dimensional data visualization classification method. First, based on the principle of multivariate pattern feature a more in-depth excavation which means that in-depth analysis of multidimensional data polyhydric FIG. Data for the dimension between 3 to 15-dimensional (small high-dimensional data), according to holographic classification (do not discard any information of a characteristic) of Thought, proposed a polyhydric FIG means and the characteristic extraction phase combined with variable integration multidimensional data visualization Category Methods. This method first radial coordinate diagram to represent multi-dimensional data, also varied diverse different types of multi-dimensional data, and then apply a single prototype graphical classifier to the automatic identification of the radial coordinate graphs. Finally, experiments to prove the effectiveness of this method. Secondly, the data for the number of dimensions in the dimension between 15 to 30 (in high-dimensional data), in order to achieve the automatic identification of the polyhydric graphics, the need to study the polyhydric FIG graphic description and conducive to plant discriminant features. To this end, this paper, a feature extraction, feature fusion and multiple graphics features primitive combination of visual classification. The first data feature extraction, enabling multidimensional data drop dimensional, not to loss of data information, the rest of the data carried vector mode long fusion, last conducted visualization, get conducive to the classification standard template, the last, in this basis on the data dimension is much larger than the 30-dimensional (high-dimensional data), this paper layered, hierarchical data by multiple shows, this method can also be extended to the dimension data classification. Finally, the application of classic dataset experiment to prove this method, the classification process, classification results visualization, and achieve a higher classification accuracy.

Related Dissertations

  1. Research on Text Classification Based on Biomimetic Pattern Recongnition,TP391.1
  2. Feature Extraction, Selection and Combination in Lipreading,TP391.41
  3. The Compression and Fusion Technique Research of Underwater Target Feature,TN911.7
  4. Research on Feature Selection and Construction in Emotion Speech Recognition,TP18
  5. An Approach for Business Process Oriented Service Dynamic Composition,TP393.09
  6. New curriculum junior high school biology \,G633.91
  7. Multilayer structure based WSN routing protocol for heterogeneous clusters,TP212.9
  8. Based on Data Distribution Characteristics of Text Classification,TP391.1
  9. The Error Analysis and Teaching of "Dao" in the Structure of "V+Dao+C",H195
  10. Research and Implementation of a Dynamic Feature Selection Method for Vehicle Recognition System,TP391.41
  11. Research on Methods of Fingerprint Image Segmentation,TP391.41
  12. Research on Face Recognition Based on AdaBoost Algorithm,TP391.41
  13. Research on Feature Extraction, Selection and Classification Algorithms for Pulmonary CAD,TP391.41
  14. Research on Distributed Fountain Codes,TN911.2
  15. Astudy on Hierarchical-Matrix-Based Algorithm for Solving Electromagnetic Problems of the Volume Integral Equation,O175.5
  16. Explore a Hierarchical Classification of Management and Preparation of "Regulatory Plan",TU981-019
  17. Identity passwords on key management,TN918.2
  18. Face recognition algorithm based on feature fusion research,TP391.41
  19. About High School Mathematics Teaching Research Triangle Module,G633.6
  20. FSVM -based data mining method and its application to intrusion detection research,TP393.08
  21. Ultra-high definition video coding based on synchronous hierarchical motion estimation architecture design,TN919.81

CLC: > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer applications > Information processing (information processing) > Pattern Recognition and devices > Image recognition device
© 2012 www.DissertationTopic.Net  Mobile