Neural Networks

What you'll learn

The key areas this subject covers, mapped to the official curriculum.

Course units

The structured path through this subject.

1. 1

   Introduction to Neural Network

   What a neural network is, the biological neuron that inspired it, the artificial neuron, and the graph view that unifies them all.

   4 topics · 4 lessons available

2. 2

   Rosenblatt’s Perceptron

   The perceptron, its learning rule, the convergence theorem that guarantees learning, and the linear separability limit that nearly killed it.

   4 topics · 4 lessons available

3. 3

   Model Building through Regression

   Building models through regression: linear models, maximum likelihood and least squares estimation, and how to judge how good the model is.

   4 topics · 4 lessons available

4. 4

   The Least-Mean-Square Algorithm

   Adaptive filtering: the Wiener filter, the least-mean-square algorithm, why it converges, and where it is used.

   4 topics · 4 lessons available

5. 5

   Multilayer Perceptron

   The multilayer perceptron: architecture, the backpropagation algorithm, why hidden layers matter, and how to keep training from collapsing into overfitting.

   4 topics · 4 lessons available

6. 6

   Kernel Methods and Radial-Basis Function Networks

   Kernel methods, radial-basis-function networks, the RBF learning process, and how RBF and MLP networks compare.

   4 topics · 4 lessons available

7. 7

   Self-Organizing Maps

   Unsupervised learning through self-organizing maps: the Kohonen map, its learning algorithm, and its real-world applications.

   4 topics · 4 lessons available

8. 8

   Dynamic Driven Recurrent Networks

   Networks with memory: feedforward versus recurrent dynamics, common recurrent architectures, backpropagation through time, and what recurrent nets are good at.

   4 topics · 4 lessons available