# Machine Learning in R
This is a repository for R scripts developed during my machine learning studies. Some of the codes have been adapted and converted into R from their original Matlab implementations.
## Classification ##
- Euclidean (euclidean_classifier)
- Mahalanobis (mahalanobis_classifier)
- Perceptron (perceptron_classifier)
- Online Perceptron (online_perceptron_classifier)
- Sum-Squared Error (sse_classifier)
## Regression ##
- plot data and (regression_plot)
- plot regression decision boundary (regression_boundary)
- generic regression wrapper function (regression_optimize)
### Linear regression ###
- Linear regression cost function and gradient (lr_cost)
- Linear regression gradient descent (lr_gradientdescent)
### Logistic regression ###
- Logistic regression cost function and gradient (logr_cost)
- Logistic regression optimizer (logr_optimize)
- Predictions (logr_predict)
### Softmax regression ###
- Softmax regression cost function and gradient (softmax_cost)
- Softmax regression optimizer (softmax_optimize)
- Predictions (softmax_predict)
### Softplus regression ###
- Softplus regression cost function and gradient (softplus_cost)
- Softplus regression optimizer (softplus_optimize)
- Predictions (softplus_predict)
## Neural Network ##
- sigmoid activation and derivative (nnet_sigmoid, nnet_dsigmoid)
- softmax activation for multi-class classification/prediction (nnet_softmax)
- forward and backward propagation (nnet_forward, nnet_backprop)
- training via gradient descent (nnet_train)
- training using stochastic gradient descent without regularization (nnet_stochastic)
- multi-class classification/prediction (nnet_predict)
- cost function for use with minimization algorithms (nnet_cost)
- fast gradient descent computation (fmincg)
- training using fast gradient descent computation (nnet_optimize)
- training using R's optim function (nnet_minimize)
## Support Vector Machines ##
- training, prediction (svm_train, svm_predict)
- decision boundary visualization (svm_plot, svm_boundary)
- linear, gaussian/rbf/erbf, polynomial kernels, sigmoid, fourier (svm_linear, svm_gaussian, svm_erbf, svm_polynomial, svm_sigmoid, svm_fourier)
- sequential minimal optimization using various kernel functions (smo2, CalcKernel)
- sequential minimal optimization boundary plotter (svcplot)
### Supported Kernels in SMO (CalcKernel) ###
- linear
- polynomial (poly)
- radial basis function (rbf)
- extended radial basis function (erbf)
- sigmoid
- fourier
- spline
- curved spline, ANOVA (curvedspline, anova)
- B-spline (bspline)
- ANOVA spline 1,2,3 (anovaspline1, anovaspline2, anovaspline3)
- ANOVA B-spline (anovabspline)
## Features ##
- normalize all features of X[m, n]: m-examples with n-features (featureNormalize)
## Principal Component Analysis ##
- compute principal components (pca)
- create projections of X (onto U) using K principal components (pca_project)
- create approximations of X using K principal components (pca_estimate)
## Anomaly Detection ##
- estimate Guassian distribution parameters: mean, variance (estimate_gaussian)
- compute multivariate Gaussian probability distribution function (multivariate_gaussian)
- determine best prediction threshold to use (select_threshold)
## Collaborative Filtering ##
- compute cost function and gradients with Regularization (cf_costfunction )
- optimize using fast gradient descent optimizer fmincg (cf_optimize, cf_fmincg_cost)
## Clustering ##
- basic sequential algorithmic scheme (BSAS)
- generalized agglomerative scheme with single and complete linking (agglom)
- run BSAS, GAS tests (test_clustering)
### k-means clustering ###
- initialization of K-centroids (kmeans_initialize)
- cluster assignment to nearest centroid (kmeans_assign)
- compute new centroids from all points belonging to K-clusters (kmeans_compute)
- run K-means algorithm for a maximum number of iterations (kmeans_run)
没有合适的资源?快使用搜索试试~ 我知道了~
主成分回归代码matlab及例子-machine-learning-r:R中的机器学习

共32个文件
r:17个
txt:12个
gitignore:1个

需积分: 31 313 浏览量
2021-05-26
12:51:25
上传
评论
收藏 85KB ZIP 举报
主成分回归代码matlab及示例R中的机器学习 这是我在机器学习期间开发的R脚本的存储库。 一些代码已从其原始Matlab实现中进行了改编并转换为R。 分类 欧几里得(euclidean_classifier) Mahalanobis(mahalanobis_classifier) 感知器(perceptron_classifier) 在线感知器(online_perceptron_classifier) Sum-Squared错误(sse_classifier) 回归 绘制数据和(regression_plot) 绘制回归决策边界(regression_boundary) 通用回归包装函数(regression_optimize) 线性回归 线性回归成本函数和梯度(lr_cost) 线性回归梯度下降(lr_gradientdescent) 逻辑回归 Logistic回归成本函数和梯度(logr_cost) 逻辑回归优化器(logr_optimize) 预测(logr_predict) Softmax回归 Softmax回归成本函数和梯度(softmax_cost) Softmax回归
资源详情
资源评论
资源推荐
收起资源包目录












































共 32 条
- 1





















weixin_38713009
- 粉丝: 8
- 资源: 920

上传资源 快速赚钱
我的内容管理 收起
我的资源 快来上传第一个资源
我的收益
登录查看自己的收益我的积分 登录查看自己的积分
我的C币 登录后查看C币余额
我的收藏
我的下载
下载帮助

会员权益专享
安全验证
文档复制为VIP权益,开通VIP直接复制

评论0