done stuff

assignment_1/report_Maggioni_Claudio.md

@@ -200,10 +200,21 @@ Comment and compare how the (a.) training error, (b.) test error and
     The classification problem in the graph, according to the data points
     shown, is quite similar to the XOR or ex-or problem. Since in 1969 that
     problem was proved impossible to solve by a perceptron model by Minsky and
-    Papert, then the
+    Papert, then that would be quite a motivation in front of my boss.
 
-1.  **Would you expect to have better luck with a neural network with
+    On a morev general (and more serious) note, the perceptron model would be
+    unable to solve the problem in the picture since a perceptron can solve only
+    linearly-separable classification problems, and even through a simple
+    graphical argument we would be unable to find a line able able to separate
+    yellow and purple dots w.r.t. a decent approximation simply due to the way
+    the dots are positioned.
+
+2.  **Would you expect to have better luck with a neural network with
     activation function $h(x) = - x \cdot e^{-2}$ for the hidden units?**
 
-1.  **What are the main differences and similarities between the
+    Boh
+
+3.  **What are the main differences and similarities between the
     perceptron and the logistic regression neuron?**
+
+

assignment_1/src/build_models.py

@@ -23,7 +23,6 @@ points = 2000
 lr = LinearRegression(fit_intercept=False)
 
 # Build x feature vector with columns for theta_3 and theta_4
-# variable name explained here: https://vimeo.com/380021022
 X = np.zeros([points, 5])
 X[:, 0] = 1
 X[:, 1:3] = xs