Radial basis function (RBF) networks have been widely used in a variety of applications, including supervised classification. Two issues are often encountered in the applications. First, the number of hidden nodes has...
详细信息
ISBN:
(纸本)9781538680988;9781538680971
Radial basis function (RBF) networks have been widely used in a variety of applications, including supervised classification. Two issues are often encountered in the applications. First, the number of hidden nodes has to be decided. Second, the settings of the basis functions have to be set. In this paper, we propose a novel RBF network approach for supervised classification applications. Given a set of training patterns, the number of hidden nodes in the hidden layer is determined by applying a self-constructing clustering algorithm on the patterns. Normalized Gaussian functions are taken to be basis functions, and their centers and deviations are set according to the clusters obtained from the clustering algorithm. The optimal values for the weights associated with the output layer are derived by adapting them to the training patterns. Experimental results are shown to demonstrate the effectiveness of the proposed approach.
This paper presents the author clustering problem and compares it to related authorship attribution questions. The proposed model is based on a distance measure called SPATIUM derived from the Canberra measure (weight...
详细信息
ISBN:
(纸本)9781538638613
This paper presents the author clustering problem and compares it to related authorship attribution questions. The proposed model is based on a distance measure called SPATIUM derived from the Canberra measure (weighted version of L-1 norm). The selected features consist of the 200 most frequent words and punctuation symbols. An evaluation methodology is presented and the test collections are extracted from the PAN CLEF 2016 evaluation campaign. In addition to those, we also consider two additional corpora reflecting the literature domain more closely. Based on four different languages, the evaluation measures demonstrate a high precision and F1 for all 20 test collections. A more detailed analysis provides reasons explaining some of the failures of the SPATIUM model.
Auction-based service provisioning and resource allocation have demonstrated strong potential in Cloud-RAN wireless network architecture and heterogeneous networks for effective resource sharing. One major technical c...
详细信息
ISBN:
(纸本)9781509050192
Auction-based service provisioning and resource allocation have demonstrated strong potential in Cloud-RAN wireless network architecture and heterogeneous networks for effective resource sharing. One major technical challenge is the integration of interference constraints in auction-based solutions. In this work we transform the interference constraint requirement into a set of linear constraints on each cluster. We tackle the generally NP-hard clustering problem by developing a novel practical suboptimal solution that can meet our design requirement. Our novel algorithm utilizes the properties of chordal graphs and applies Lexicographic Breadth First Search (Lex-BFS) algorithm for cluster splitting. This polynomial time approximate algorithm searches for maximal cliques in a graph by generating strong performance in terms of subgraph density and probability of optimal clustering without suffering from the high complexity of the optimal solution.
The article presents a particular comparison of selected clustering algorithms of data obtained by interferometric methods using artificial neural networks. For the purposes of the experiment original data from Szczec...
详细信息
ISBN:
(纸本)9781509025183
The article presents a particular comparison of selected clustering algorithms of data obtained by interferometric methods using artificial neural networks. For the purposes of the experiment original data from Szczecin Port have been tested. For collecting data authors used the interferometric sonar system GeoSwath Plus 250 kHz. GeoSwath Plus offers very efficient simultaneous swath bathymetry and side scan seabed mapping. During the use of Kohonen's algorithm, the network, during learning, use the Winner Take All rule and Winner Take Most rule. The parameters of the tested algorithms were maintained at the level of default. During the research several populations were generated with number of clusters equal 9 for data gathered from the area of 100m(2). In the subsequent step statistics were calculated and outcomes were shown as spatial visualization and in tabular form.
With the development of economy and technology,introducing and training talents have become the key driving force in the world which can enhance the competitive strength of the whole ***,the strategies of strengthenin...
详细信息
With the development of economy and technology,introducing and training talents have become the key driving force in the world which can enhance the competitive strength of the whole ***,the strategies of strengthening the universities and colleges with more talented people and making efforts to implement the construction of "Double top" are put forward in the same *** of clustering analysis have been widely used in the actual *** this study,an effective clustering analysis model by comparing the clustering analysis under different dimensionality reduction methods is ***,preprocess the data about talent introduction which is collected from Zhejiang University of Finance and Economics,and use Principal Component Analysis(PCA),Weighted Principal Component Analysis(Weighted-PCA) and Random Forest(RF) to reduce the dimensions of the ***,use K-means clustering algorithm and K-medoids clustering algorithm to cluster the preprocessed *** classification results indicate that the K-medoids algorithm with Weighted-PCA is superior to other clustering methods in this illustrative *** addition,the experiment divides talents into high-end talents and mid-end *** looking into the analysis of the characteristics of the clustering results,some targeted advices on the talents introduction in colleges can be provided.
clustering by fast search and find of density peaks（CFSFDP） is honored by its simplicity and speed. However, the hyper-parameter dc can only be determined through empirical experience, and for those datasets contain...
详细信息
clustering by fast search and find of density peaks（CFSFDP） is honored by its simplicity and speed. However, the hyper-parameter dc can only be determined through empirical experience, and for those datasets containing points that are closer to other cluster center, the algorithm does not perform well. In this work, we proposed a new method called Improved CFSFDP（I-CFSFDP）, which makes two modifications compared to CFSFDP. Firstly, a new indicator for density is introduced to eliminate the effect of dc on clustering results. Secondly, cluster diffusion model was proposed to cluster remaining points after finding cluster centers. When regarding datasets as graphs, this process can be abstracted as finding the minimum spanning forest model in a graph, and each spanning tree represents a cluster in the dataset. I-CFSFDP is comprehensively evaluated on several datasets with arbitrary distribution and has demonstrated that I-CFSFDP is distinctly more accurate and robust than CFSFDP and DBSCAN.
At present,we assess the air quality levels based on AQI which requires complex calculations and cannot evaluate the air quality comprehensively and *** air quality evaluation method not only can cluster air quality d...
详细信息
At present,we assess the air quality levels based on AQI which requires complex calculations and cannot evaluate the air quality comprehensively and *** air quality evaluation method not only can cluster air quality data precisely,but also can identify the air quality using unsupervised learning without any prior *** collect 2600 days' of air pollutants data from baoding,suzhou and sanya which are clustered into 11 categories.A comparison between our clustering results and the air quality levels according to the traditional algorithm has been ***,we can identify a new air quality data set with the accuracy of about 96.5%based on variational autoencoder(VAE) *** research will help people assess the air quality with air pollutants indexes more conveniently and scientificly.
In sensor network applications, energy is considered to be the most critical constraint. Our aim is to reduce the energy consumption as well as gain efficient accuracy by integrating the concept of tracking algorithm ...
详细信息
In sensor network applications, energy is considered to be the most critical constraint. Our aim is to reduce the energy consumption as well as gain efficient accuracy by integrating the concept of tracking algorithm with heuristic approach. An energy efficient prediction-based clustering algorithm is being proposed here, based on the fact that the movements of the tracked objects are sometimes predictable. The proposed algorithm helps us to reduce the transmission distance between transmitter and receiver nodes and decreases the number of transmitted packets. Only the special nodes are being activated and rest of the nodes goes to sleep mode for energy saving. The proposed method has better performance than other prediction based methods in energy efficiency when simulated
Using clustering algorithm to improve the effectiveness of test case prioritization has been well recognized by many researchers. Software fault prediction has been one of the active parts of software engineering, but...
详细信息
ISBN:
(纸本)9781509055074
Using clustering algorithm to improve the effectiveness of test case prioritization has been well recognized by many researchers. Software fault prediction has been one of the active parts of software engineering, but to date, there are few test cases prioritization technique using fault prediction. We conjecture that if the code has a fault-proneness, the test cases covering the code will find fault with higher probability. In addition, most of the existing test cases prioritization techniques using clustering algorithm don't consider the number of clusters. Thus, in this paper, we design a test case prioritization based on clustering approach combining fault prediction. We consider the method to obtain the best number of clusters and the clustering prioritization based on the results of fault prediction. To investigate the effectiveness of our approach, we perform an empirical study using an object which contains test cases and faults. The experiment results indicate that our techniques can improve the effectiveness of test case prioritization.
Data classification is one of the core technologies in the field of pattern recognition and machine learning, which is of great theoretical significance and application value. With the increasing improvement of data a...
详细信息
ISBN:
(数字)9783319598581
ISBN:
(纸本)9783319598581;9783319598574
Data classification is one of the core technologies in the field of pattern recognition and machine learning, which is of great theoretical significance and application value. With the increasing improvement of data acquisition, storage, transmission means and the amount of data, how to extract the essential attribute data from massive data, data accurate classification has become an important research topic. Inverse nth n order gravitational field is essentially a generalization of the n order in the physics, which can effectively describe the interaction between all the particles in the gravitational field. This paper proposes a new inverse nth power gravitation (I-n-PG) based clustering method is proposed for data classification. Some randomly generated data samples as well as some well-known classification data sets are used for the verification of the proposed I-n-PG classifier. The experiments show that our proposed I-n-PG classifier performs very well on both of these two test sets.
暂无评论