MIT xPRO: DSx Data Science and Big Data Analytics: Making Data-Driven Decisions - Case Study 3.2

main.py

import matplotlib.pyplot as plt
import numpy as np


def plot_dataset(x, y, legend_loc='lower left'):
	fig, ax = plt.subplots()

	ax.scatter(
		x[y == 1, 0],
		x[y == 1, 1],
		c='r',
		s=100,
		alpha=0.7,
		marker='*',
		label='Sea Bass',
		linewidth=0)

	ax.scatter(
		x[y == - 1, 0],
		x[y == -1, 1],
		c='b',
		s=100,
		alpha=0.7,
		marker='o',
		label='Salmon',
		linewidth=0)

	ax.axhline(
		y=0,
		color='k')

	ax.axvline(
		x=0,
		color='k')

	ax.set_xlabel('Length')
	ax.set_ylabel('Lightness') 
	ax.set_aspect('equal')

	if legend_loc:
		ax.legend(loc=legend_loc, fancybox=True).get_frame().set_alpha(0.5)
		ax.grid('on')

	plt.show()


# For the three figures in part 1
x = np.array([
	[2, 1], 
	[0, -1], 
	[1.5, 0], 
	[0, 1], 
	[-1, 1], 
	[-3, 0],
	[1, -1],
	[2, - 1],
	[3, -2],
	[3, 1],
	[-2, 1.5],
	[-3, 0.5],
	[-1, 2]]) 

y = np.array([1, 1, 1, -1, -1, -1, 1, 1, 1, 1, -1, - 1, -1]) 

# Show Part 1 - Figure 1
plot_dataset(x, y)

x2 = np.vstack([
	x, 
	np.array([0, -0.2])])

y2 = np.hstack([y, np.array([-1])]) 

# Show Part 1 - Figure 2
plot_dataset(x2, y2)

x3 = np.array([
	[4, 1],
	[-2, 1],
	[-2, - 4],
	[-1, -1],
	[2, -1],
	[-1, -3],
	[3, 2],
	[1, 2.5],
	[-3, -1],
	[-3, 3],
	[0,	-2],
	[4, -2],
	[3, -4]])

y3 = np.array([1, 1, 1, -1, - 1, -1, 1, 1, 1, 1, -1, -1, -1])

# Show Part 1 - Figure 3
plot_dataset(x3, y3, legend_loc='lower right')

# For the sigmoid network in part 2 
def sigmoid(inputs):
	return 1.0 / (1.0 + np.exp(-inputs))


def nn_2layer(inputs):
	return np.sign(sigmoid(inputs[:, 0]) + sigmoid(-inputs[:, 1]) - 1.5)


def plot_decision_boundary(network):
	x0v, x1v = np.meshgrid(np.linspace(-2, 8, 20), np.linspace(-8, 2, 20)) 
	x4 = np.hstack([x0v.reshape((-1, 1)), x1v.reshape((-1, 1))])
	y4 = network(x4)
	plot_dataset(x4, y4, legend_loc=None)


# Show Part 2 - Figure 1
plot_decision_boundary(nn_2layer)

# For the ReLU network in Part 2 
def relu(inputs):
	return np.maximum(0, inputs)


def nn_2layer_relu(inputs):
	return np.sign(relu(-inputs[:, 0]) + relu(inputs[:, 1]) - 0.1)


# Show Part 2 - Figure 2
plot_decision_boundary(nn_2layer_relu)

requirements.txt

matplotlib==3.3.4
numpy==1.18.0