Neural network for machine learning

Realize the projects of coursera course "Neural Networks for Machine Learning", which is taught by Prof. Geoffrey Hinton.

Author: Jian Wang and Shao Tang

Contribution: Write python code instead of matlab, which is the default for this course, to get the numerical results and answer questions of each project.

Programming Assignment 1: The perceptron learning algorithm

Use perceptron learning algorithm to classify four two-classes dataset.Set iteration numbers equal to 5 as examples

Reuslt for dataset1:

Reuslt for dataset2:

Reuslt for dataset3:

Reuslt for dataset4:

The left plot shows the dataset and the classification boundary given by the weights of the perceptron. The negative examples are shown as circles while the positive examples are shown as squares. If an example is colored green then it means that the example has been correctly classified by the provided weights. If it is colored red then it has been incorrectly classified. The middle plot shows the number of mistakes the perceptron algorithm has made in each iteration so far. The right plot shows the distance to some generously feasible weight vector if one has been provided (note, there can be an infinite number of these). Points that the classifier has made a mistake on are shown in red, while points that are correctly classified are shown in green.

We can see that data1, data3 can be linearly separable, while data2 and data4 can not.