Add files via upload

Linear_and_Logistic_Regression/LWR_gif.m

@@ -0,0 +1,33 @@
+clc; close all; rng(0);
+
+% Generate Sine Data Set
+N = 100;                        % No. of training points
+x = linspace(0,12,N)';          % Training data, x
+y = sin(x) + 0.5*rand(N,1);     % Training data, y
+y = (y - mean(y))./std(y);      % Standard scaling on y
+
+scatter(x,y,'b','filled','MarkerFaceAlpha',0.3); 
+grid on; box on; set(gcf,'Color','w'); 
+
+xp = linspace(-2,14,200)';      % Finely spaces x values 
+yp = zeros(size(xp));
+
+for tau = 0.1:0.02:3
+    weighting = @(x,X,tau) exp(-(x-X).^2/(2*tau^2));
+    for j = 1:length(xp)
+        W = diag(weighting(xp(j),x,tau));
+        X = [ones(N,1), x];     
+        w = (X'*W*X)\(X'*W*y);  % Locally weighted linear reg.
+        yp(j) = [1 xp(j)]*w;    % Predict on xp(j)
+    end
+
+    hold on; p = plot(xp,yp,'k','LineWidth',1.5);
+    set(gcf,"Position",[488,506.6,560,255.4]);
+    title(sprintf('LWR, tau = %.4f',tau))
+    axis([-2 14 -3 3]);
+
+    % For creating a GIF
+    % exportgraphics(gcf,'LWR_tau.gif','Append',true);
+
+    pause(0.5); delete(p);
+end