逻辑斯蒂回归实例-matlab

% 此程序计算 clear all close all clc %% 读取数据 load q1x.dat; load q1y.dat; x = [ones(size(q1x, 1), 1) q1x]; y = q1y; %% 训练 theta1 = zeros(size(x, 2), 1); eps = 1e-6; tol = 1; while tol>eps theta0 = theta1; H = Hassian(x, y, theta0); J = Derivation(x, y, theta0); theta1 = theta0 - inv(H)*J; tol = norm(theta1-theta0); end theta = theta0 %% 绘图 x1 = x(:, 2); x2 = x(:, 3); index1 = find(y);index2 = find(y == 0); plot(x1(index1), x2(index1), 'bo', x1(index2), x2(index2), 'rx'); hold on xmin = min(x1); xmax = max(x1); fun = @(x)(-theta(2)/theta(3)*x - theta(1)/theta(3)); plot([xmin, xmax], [fun(xmin), fun(xmax)], '-g');