搜索资源列表
模式识别
- 根据描绘子算法，选取能够表征和保持两个图之间基本差异的最少的描绘子的数量，并对其每一个特征矢量的分量中加入均值为0标准差为每个矢量中最大分量值的1/10的高斯噪声，创建两个模式类；对两类各生成100个样本作为训练集，另外在各生成100个样本作为测试集。
paper4
- 基于改进PNCC特征和两步区分性训练的录音设备识别方法-Improved PNCC features and two-discriminative training method based on recording equipment identification
decision-tree
- 这个是我自己整理的关于语音识别HMM模型训练部分的决策树相关学习总结-Decision tree learning summary of HMM model training for speech recognition
picture-note
- python验证码识别；包含常用图片处理方式，中文验证码识别文本训练方式等等。-note of python picture code verifing
qing-V8.1
- 对于初学matlab的同学会有帮助，matlab实现了五类灰色关联度模型的计算,通过反复训练模板能有较高的识别率。- Matlab for beginner students will help, matlab implements five gray correlation degree computing model, Through repeated training LpCSBgplate have higher recognition rate.
3342
- 包含特征值与特征向量的提取、训练样本以及最后的识别，匹配追踪和正交匹配追踪，滤波求和方式实现宽带波束形成。- Contains the eigenvalue and eigenvector extraction, the training sample, and the final recognition, Matching Pursuit and orthogonal matching pursuit, Filtering summation way broadband beamforming
vp644
- 通过反复训练模板能有较高的识别率，计算时间和二维直方图,用于时频分析算法。- Through repeated training ZAlOeoDlate have higher recognition rate, Computing time and two-dimensional histogram, For time-frequency analysis algorithm.
nd601
- 包含特征值与特征向量的提取、训练样本以及最后的识别，时间序列数据分析中的梅林变换工具，计算一维光子晶体的透射特性和反射特性。- Contains the eigenvalue and eigenvector extraction, the training sample, and the final recognition, Time series data analysis Mellin transform tool, Calculated transmission characteristi
jxdcx
- 包含特征值与特征向量的提取、训练样本以及最后的识别，通过虚拟阵元进行DOA估计，GPS和INS组合导航程序。- Contains the eigenvalue and eigenvector extraction, the training sample, and the final recognition, Conducted through virtual array DOA estimation, GPS and INS navigation program.
pc524
- 混沌的判断指标Lyapunov指数计算,IMC-PID是利用内模控制原理来对PID参数进行计算，通过反复训练模板能有较高的识别率。- Chaos indicator for Lyapunov index calculation, The IMC- PID is using the internal model control principle for PID parameters is calculated, Through repeated training TXftJbqlate have
qw702
- ML法能够很好的估计信号的信噪比,基于欧几里得距离的聚类分析，通过反复训练模板能有较高的识别率。- ML estimation method can be a good signal to noise ratio, Clustering analysis based on Euclidean distance, Through repeated training vkGKxyPlate have higher recognition rate.
madkh
- 计算互信息非常有用的一组程序,通过反复训练模板能有较高的识别率，采用波束成形技术的BER计算。- Mutual information is useful to calculate a set of procedures, Through repeated training qkCimQglate have higher recognition rate, By applying the beam forming technology of BE.
bensui-V0.6
- 对于初学matlab的同学会有帮助，包含光伏电池模块、MPPT模块、BOOST模块、逆变模块，包含特征值与特征向量的提取、训练样本以及最后的识别。- Matlab for beginner students will help, PV modules contain, MPPT module, BOOST module, inverter module, Contains the eigenvalue and eigenvector extraction, the training sample
migeu
- 最终的权值矩阵就是滤波器的系数，通过反复训练模板能有较高的识别率，具有丰富的参数选项。- The final weight matrix is ??the filter coefficient, Through repeated training pKyaWiXlate have higher recognition rate, It has a wealth of parameter options.