In order to improve the reliability of the cold chain logistics supply chain and shorten the response time of the supply chain, a quality evaluation method of agricultural cold chain logistics supply chain based on k-...
详细信息
Sports competition data analysis and strategy optimization are important ways to enhance athlete competitiveness and team collaboration. The current competition analysis and strategy formulation have strong subjectivi...
详细信息
Sports competition data analysis and strategy optimization are important ways to enhance athlete competitiveness and team collaboration. The current competition analysis and strategy formulation have strong subjectivity, making it difficult to deeply understand the performance characteristics and patterns of athletes and teams. Traditional analysis methods cannot accurately identify the performance differences of different athletes, and there are limitations in their feature recognition and classification. In order to enhance the scientificity of strategy formulation and improve the performance of athletes in competitions, this article combined the k-means clustering algorithm and focused on basketball sports to conduct an in-depth analysis of sports competition data analysis and strategy optimization. Firstly, the competition data was collected and preprocessed. Then, feature selection was carried out from three dimensions: competition results, player performance, and team characteristics. Finally, the k-means clustering algorithm was used to perform hierarchical clustering on the original data through a hierarchical method. To verify its effectiveness, this article conducted practical analysis on the data of nearly 5 basketball competitions in 10 university basketball leagues in a certain province and optimized strategies based on cluster analysis. The results showed that in terms of player performance, compared to before optimization, the average number of rebounds, assists, and steals of team players optimized based on algorithm strategy increased by about 38.9%, 25.0%, and 63.2%, respectively. The conclusion indicates that the application of k-means clustering algorithm in sports competition data analysis and strategy optimization can help improve the competitive level of athletes and enhance their performance.
In this paper, we report a hardware/software (MW/SW) co-designed k-means clustering algorithm with high flexibility and high performance for machine learning, pattern recognition and multimedia applications. The contr...
详细信息
In this paper, we report a hardware/software (MW/SW) co-designed k-means clustering algorithm with high flexibility and high performance for machine learning, pattern recognition and multimedia applications. The contributions of this work can be attributed to two aspects. The first is the hardware architecture for nearest neighbor searching, which is used to overcome the main computational cost of a k-means clustering algorithm. The second aspect is the high flexibility for different applications which comes from not only the software but also the hardware. High flexibility with respect to the number of training data samples, the dimensionality of each sample vector, the number of clusters, and the target application, is one of the major shortcomings of dedicated hardware implementations for the k-meansalgorithm. In particular, the HW/SW k-meansalgorithm is extendable to embedded systems and mobile devices. We benchmark our multi-purpose k-means system against the application of handwritten digit recognition, face recognition and image segmentation to demonstrate its excellent performance, high flexibility, fast clustering speed, short recognition time, good recognition rate and versatile functionality. (c) 2012 Elsevier B.V. All rights reserved.
It is critical to forecast the electric load for a region. Traditional electric load forecasting frequently predicts the load of multiple transformers in the region after directly summing them, but directly predicting...
详细信息
ISBN:
(纸本)9781665490542
It is critical to forecast the electric load for a region. Traditional electric load forecasting frequently predicts the load of multiple transformers in the region after directly summing them, but directly predicting the total load after accumulating the load of each distribution transformer will weaken the individual time-series characteristics and reduce prediction accuracy, so it is necessary to forecast the total load of the region while appropriately retaining the individual time-series characteristics. To address the aforementioned issues, this paper clusters transformer load curves, narrows the load characteristics of the same category of transformers using the weighting concept. It stores the load characteristics of multiple transformers in a neural network. This paper develops a virtual transformer load forecasting model in which the sum of all transformer loads in the region is a number of times the virtual transformer load forecasting result.
With the growing popularity of the network, product information filled in the many pages of the Internet, which you want to get the information you need on these pages tend to consider clustering information, and the ...
详细信息
ISBN:
(纸本)9781467365932
With the growing popularity of the network, product information filled in the many pages of the Internet, which you want to get the information you need on these pages tend to consider clustering information, and the current explosive growth of data so that the information mass storage condition occurs, clustering to facing the problems such as large calculation complexity and time consuming, then the traditional k-means clustering algorithm does not meet the needs of large data environments today, so this article combined with the advantages of the Hadoop platform and MapReduce programming model is proposed the k-means clustering algorithm for large-scale chinese commodity information Web based on Hadoop. Map function calculates the distance from the cluster center for each sample and mark to their category, Reduce function intermediate results are summarized and calculated new clustering center for the next round of iteration. Experimental results show that this method can better improve the clustering processing speed.
Spectral ratio methods have been widely used in evaluation of nonlinear seismic site response. Nevertheless, it remains inefficient and subjective to identify stations with nonlinear site response according to empiric...
详细信息
Spectral ratio methods have been widely used in evaluation of nonlinear seismic site response. Nevertheless, it remains inefficient and subjective to identify stations with nonlinear site response according to empirical threshold values of spectral ratio nonlinear degree indicators. This study, which was the first to apply the machine learning clusteringalgorithm to address this problem, used the September 6, 2018 M(w)6.6 Hokkaido Iburi-Tobu earthquake (Japan) as an example. First, we calculated the surface/borehole and horizontal/vertical spectral ratios using strong ground motion data recorded by kik-net vertical array and k-NET stations, respectively. The degree of nonlinear site response (DNL) and percentage of nonlinear site response (PNL) were computed using the difference between the strong motion of the mainshock and weak aftershocks as the reference for linear site response. Then, the k-means clustering algorithm was incorporated in the identification of nonlinear site response using the DNL, PNL, strength of ground motion (PGA) and site condition (V-S20 or V-S30) as explanatory variables. After careful multicollinearity diagnosis and confirmation of the optimum clustering number, we successfully classified the stations into two clusters with nonlinear and linear site responses. Overall, the clustering results were found in good agreement with the classification results based on empirical thresholds of several nonlinear indicators. For the stations identified with nonlinear site response, the reduction of amplification and frequency shift could be observed from the spectral ratio curves regarding the ground motions in the mainshock and the reference weak aftershocks, demonstrating typical nonlinearity response characteristics. Furthermore, a comprehensive indicator of nonlinear site response occurrence probability (NLscore) was obtained from a linear weighted combination of the normalized variables (PGA, V-S30/V-S20, DNL and PNL). The NLscore ranking of th
For apple harvesting robot, it is difficult to acquire the coordinates of occluded apples accurately in natural scenes, which is important in implementing picking tasks. In this paper, a method on automatic recognitio...
详细信息
For apple harvesting robot, it is difficult to acquire the coordinates of occluded apples accurately in natural scenes, which is important in implementing picking tasks. In this paper, a method on automatic recognition and localization of occluded apples was proposed. Firstly, an apple recognition algorithm based on k-meansclustering theory was described. Secondly, convex hull information which was obtained by following the contours of extracted apple regions was used to extract the real apple edges. Finally, three points from these real edges were selected to estimate the centers and radius of apples. This algorithm was tested and compared with traditional Hough transform method (HT method) and contour curvature method (CC method) and 125 apple images were used to test the effectiveness of these methods. Four parameters including Segmentation Error (SE), False Positive Rate (FPR), False Negative Rate (FNR) and Overlap Index (OI) were used to evaluate the performance of these methods. Experimental results showed that SE of the presented method was decreased by 14.399 and 30.782 % when compared to CC method and HT method respectively, FPR by 7.234 and 11.728 % and OI was increased by 18.644 and 30.938 %. FNR of the proposed method was 0.912 % lower than CC method, while it was 5.869 % higher than HT method. The experimental results indicated that the proposed method could get much better localization rate than Hough transform method and contour curvature method, thus it could be concluded that the algorithm is an efficient means for the recognition and localization of occluded apples.
Computational fluid dynamics (CFD) modelling is a scientific tool to provide fluid dynamics and chemical simulation that facilitates understanding of the complex combustion phenomenon in engine studies. With the advan...
详细信息
Computational fluid dynamics (CFD) modelling is a scientific tool to provide fluid dynamics and chemical simulation that facilitates understanding of the complex combustion phenomenon in engine studies. With the advance of Machine Learning (ML) technology, the big data from CFD results can be intelligently recognized and classified, thus ease the data post-processing. This study proposed an integrated analysis that uses CFD simulation results of scalar distributions and k-means clustering algorithm to optimally partition engine combustion chamber into different zones. Therefore, the space of combustion chamber was automatically divided into light soot zones and heavy soot zones based on the clustering results on local equivalence ratio (ER) and temperature. Consequently, the surveys of soot mitigation by Reactivity Controlled Compression Ignition (RCCI) engines combustion mode were carried out as well as corresponding sooting tendency by CFD numerical study. The localized soot depositions in each zone under varied combustion boundaries were compared, hence improving the development of control strategy with numerical modellings and machine learning techniques.
Firstly, this paper introduces the types of clusteringalgorithm, and introduces the classical k-meansalgorithm and canopy algorithm in detail. Then, combining the map reduce computing model and spark cloud computing...
详细信息
Firstly, this paper introduces the types of clusteringalgorithm, and introduces the classical k-meansalgorithm and canopy algorithm in detail. Then, combining the map reduce computing model and spark cloud computing framework, this paper introduces the parallel Canopy-k-meansalgorithm after using Canopy algorithm to optimize the initial value of k-meansalgorithm. However, because Canopy algorithm needs to introduce a new distance threshold parameter T2, and the parameter needs to be set by human experience, it is difficult to determine the parameter artificially for large data, so this paper proposes a parallel adaptive Canopy-k-meansalgorithm, which can be used in cloud computing framework to determine the distance threshold parameter T2 adaptively based on statistical method. Using the parallelism of Map-Reduce computing model, the parallel Canopy-k-meansalgorithm is optimized by adaptive parameter estimation, which solves the problem that parameters depend on manual experience selection in Canopy process. After introducing the relevant theories and derivation process of this algorithm, cloud computing experiment platform is built based on the Spark framework, and the contrast experiments were performed using the Stanford Large Network Dataset Collection (SNAP) dataset and self-built Dimension Networks dataset. The experimental results show that the proposed method is effective.
This study describes the performance analysis of the non-homogeneity detector (NHD) with various normalisation methods for the space-time adaptive processing (STAP) of airborne radar signals under the non-homogeneous ...
详细信息
This study describes the performance analysis of the non-homogeneity detector (NHD) with various normalisation methods for the space-time adaptive processing (STAP) of airborne radar signals under the non-homogeneous clutter environments. The authors can calculate a threshold value from the statistical analysis of generalised inner product (GIP) using the normalisation method using mean, median and the k-means clustering algorithm of training data snapshots in the NHD process. The selected homogeneous data using the threshold value are used to recalculate covariance matrix of the total interference. To evaluate the performance of the covariance matrix, the authors calculated the eigenspectra and signal to interference noise ratio (SINR) loss. The accuracy of the recalculated covariance matrix is verified by the modified sample matrix inversion (MSMI) test statistic for the target detection. Projection statistics (PS) based on GIP is also used to compare the performance of detecting single and multiple targets. The authors' simulation results demonstrate that the k-means clustering algorithm as a normalisation method for both GIP and GIP-based PS can improve the STAP performance in the severe non-homogeneous clutter environment even under the multiple targets scenarios, compared to the other normalisation methods.
暂无评论