共查询到20条相似文献,搜索用时 46 毫秒
1.
在数据挖掘中,决策树方法是一个重点研究方向。ID3方法是著名的决策树算法,在这种算法中,互信息的计算依赖于特征值数目较多的特征,这样不太合理.为此,从离散度的角度,对ID3算法进行改进,通过2种算法的对比实验,证明利用改进后的算法挖掘分类规则,不仅提高了分类的正确率,而且非常高效。 相似文献
2.
基于决策树的一种改进算法 总被引:2,自引:0,他引:2
首先介绍了ID3算法的基本思想,然后讨论了决策树算法中的难点问题,针对ID3算法中所存在的不足,提出了一种利用优值法的思想来改进信息增益的算法,并且与ID3算法进行了实验对比。实验表明,这种方法从树的规模和分类精度都优于许多决策树算法,使决策效率明显提高。 相似文献
3.
ID3是决策树学习的核心算法.本文提出了基于ID3算法对多类样本分类的一种快速分类的算法.提高了计算机的分类速度。 相似文献
4.
5.
6.
针对ID3算法倾向于取值较多的属性的特点,同时为避免特定领域主观属性影响分类结果,提出了一种改进的决策树算法。该方法利用决策树算法构造决策树,通过对分类结果中主客观属性进行标记并逻辑运算,最终得到较客观的决策信息,并进行实验验证。 相似文献
7.
8.
为了更好地提高入侵检测的准确率,节省检测时间,文章提出了一种基于增量式的决策树检测算法。该方法采用了基于粒度决策熵和改进的主成分分析方法对数据集中的冗余以及不相关属性进行归类、降维。该方法将数据挖掘增量学习技术与决策树分类算法相结合,在属性降维后的决策树基础上,对于新的测试样本实例,引入扩展贝叶斯结点,比较贝叶斯分类方法与决策树分类方法的准确率,返回更新后的决策树。针对属性降维,主成分分析方法在约简属性的基础上,能够有攻击分类结果准确率高、耗时少的特点。将增量式决策树算法与贝叶斯算法、ID3算法进行对比,发现在检测精确率、检测效率的指标下,增量式决策树算法在一定程度上优于其他算法,并且误报率有效降低。 相似文献
9.
10.
为有效定位识别和提取网络流量序列的暂态性异常特征,针对网络异常流量特征扰动性和暂态性特点,提出一种基于小波分解的二叉分类回归决策树主分量特征优化跟踪特征提取算法。利用训练集建立决策树模型,采用二叉分类回归决策树模型进行主分量特征优化跟踪建模,利用双正交提升小波分解得到的各层细节信号对暂态性扰动特征的敏感性,通过小波分解得到各层细节信号,将提取的小波分层细节信号的奇异值分解特征再返回到决策树主分量特征优化跟踪模型中,实现网络流量异常特征的定位提取和识别。仿真实验表明,改进算法的抗干扰能力和分辨率提高显著,暂态性异常特征谱图分辨能力提高,异常特征分布谱清晰可见,展示了较好的特征提取和状态识别性能。 相似文献
11.
12.
高亮度显示器对黑色字符的显示对比度很难做到精确的测定。文中介绍了几种解决这一问题的方案。 相似文献
13.
14.
为了银河系中心远红外的球载观测,采用了 Ge-Bolometer(锗测辐射热探测器)作为天体远红外辐射的探测器,由于气球的飞行高度在35km 附近,所以在球上的液氦致冷探测器的工作温度在1.7K 左右。本文讨论了在实验室中解决这样低温的致冷方法,最后讨论了探测器性能参数的测试装置和测试结果。 相似文献
15.
16.
非傍轴厄米-正弦-高斯光束的特性 总被引:1,自引:0,他引:1
为了研究厄米-正弦-高斯(HSiG)光束在非傍轴情形下受参量影响的特性,基于光强二阶矩理论,经过一系列复杂的数学计算,得出了非傍轴厄米-正弦-高斯光束的远场发散角、束腰宽度以及光束传输因子的解析式,并通过数值模拟,得到了它们与离心参量的关系。结果表明,在束腰宽度与波长比取值逐渐减小时,随着离心参量趋于0,当阶数为奇数和偶数时.发散角分别趋于63.435°和73.898°。束腰宽度和光束传输因子随参量的增加呈现较大的起伏,非傍轴情形下的光束传输因子取值与傍轴情况有很大的不同,其受参量影响不仅可以小于1,甚至趋于0。 相似文献
17.
Chien-Cheng Lee Cheng-Yuan Shih Chih-Chang Yu Wei-Ru Lai Bor-Shenn Jeng 《Journal of Signal Processing Systems》2011,64(3):291-303
This paper proposes a vision-based fingertip handwriting recognition system to provide an alternative to input devices. Traditional
handwriting recognition systems are limited because they require a specific or expensive input device, such as pen, tablet,
or touch panel. Recently, cameras have gradually become standard components in many computer-based products. Therefore, a
fingertip and camera combination provides a flexible and convenient input device. The proposed system combines fingertip detection,
trajectory feature extraction, and character recognition. First, fingertip moving trajectories are tracked and recoded. The
proposed cyclic chain code histograms are then obtained from the trajectories and used as features in the following recognition
process. An improved radial basis function (RBF) neural network is used to recognize handwritten characters. Experimental
results show that the proposed novel input system is feasible and effective. This study also presents possible applications
for camera input devices. 相似文献
18.
Character fonts on raster scanned display devices (e.g., the home television receiver) are usually represented by arrays of bits and displayed as a matrix of black and white (or color) dots. In VIDEOTEX systems, due to the limitations on bandwidth, number of scanning lines, and the interlace of the broadcast television standard, each frame is allowed to contain about 21 rows of 40 characters each; each character occupying a rectangular area of 12 picture elements by 20 lines. The character sets that have traditionally been used in VIDEOTEX systems, have not used the multilevel display capabilities of the television receiver. Due to interlace, this results in flicker at every sharp color vertical transition. In this note, we show how the multilevel capabilities of a television receiver can be used by nonlinearly filtering the characters generated by a conventional VIDEOTEX character generator such that each vertical transition is made less sharp. Although the characters are more blurred due to filtering, the apparent resolution of the display is increased allowing smaller fonts to be used. Flicker is reduced to a large extent, thus, decreasing the fatigue due to prolonged viewing of the display. A font occupying 9X16 rectangular area per character is proposed. It results in about 65 percent more characters per 'page-screen' compared to the present VIDEOTEX system. The hardware required for such nonlinear filtering is minor, since, storage of multilevel intensities of the characters is not required due to local processing at the receiver. 相似文献