
GRADED EXERCISE 2 


Graded Exercise 2  Solutions 


Grades 
1st Graded Exercise  Results 
The results of the first graded session which took place on 2.4.2019. The total number of points is 15. 

Graded Exercise 1 Solutions 
The solutions of the first graded exercise 

2nd Graded Exercise  Results File 


Introduction 
Course Introduction 


Naive Bayes 


Images as vectors 


Python and NumPy Primer 


Exercise Session 1 


Solutions 


Nearest Neighbors 
K nearest neighbors 


Bishop: 2.5.2 Nearestneighbour methods 


Exercise Session 2 


Solutions (jupyter notebook) 


Solutions (html) 


Linear Classification 
Logistic regression 


Bishop: 4.1.7 The Perceptron Algorithm 


Bishop: 4.3.2 Logistic regression 


Optimization 


Triggs: Bundle Adjustment — A Modern Synthesis 


Exercise Session 3 


Solutions (jupyter notebook) 


Solutions (html) 


Exercise Session 4 


Solutions (jupyter notebook) 


Solutions (html) 


From Linear to NonLinear Classification 
HighDimensional Features and Kernels 


Proof of Cover's theorem 


Bishop: 7.1 Maximum margin classifiers 


Support Vector Machine vs Logistic Regression 


Boosting 


Bishop: 14.3 Boosting 


Trees and Forests 


Bishop: 14.4 Treebased models 


Exercise Session 5 


Solution(jupyter notebook) 


Solution(html) 


Exercise Session 6 


Solution(notebook) 


Solution(html) 


Neural Nets 
Multilayer perceptrons 


Bishop: 5 Neural networks 


Nielsen: Deep learning tutorial 


Kingma: Adaptive Moment Estimation 


Telgarsky: Benefits of depth 
Telgarsky paper showing the benefits of deeper networks. 

Convolutional networks 
Convolutional networks designed to handle images. 

AlphaGo 


Exercise Session 7 


solution (notebook) 


solution (html) 


Exercise Session 8 


Exercise Session 8  WINDOWS 


solution (html) 


Solutions (notebook) 


Regression 
Linear regression and least squares 


Bishop: 3.1 Linear basis function models 


Nonlinear least squares 


Fua: Constrained least squares 


Gaussian processes 


Bishop: 6.4 Gaussian processes 


Regression Trees 


Caruana: An empirical evaluation of supervised learning in high dimensions 


Neural Nets 
Neural networks and Autoencoders 

Exercise Session 9 


Exercise 9 Solution (HTML) 


Exercise 9 Solution 


Exercise session 10 


Exercise 10 Solution 


Modeling over time 
Recurrent networks 


Human Brains 
Brain networks 


Gerstner: Spiking Neuron Models 


Summary 
Conclusion 
