Learning and Intelligent Systems

The course will introduce the foundations of learning and making predictions from data. We will study basic concepts such as trading goodness of fit and model complexity. We will discuss important machine learning algorithms used in practice, and provide hands-on experience in a course project. VVZ Information is available here.

---

News

• The final exam reviews will take place on October 4 and October 5 at 3pm in CAB G 66.2
• The final exam will take place on August 13, from 9:00-11:00 (120 minutes) in HIL F15 (Hönggerberg).
• The exam review session will be held on Thursday, August 4, from 13-15 in ML D 28.
• There is a typo in the slides of Tue 22.3. “Sparsity (Lasso, L1-SVM), Class Imbalance”: it should say argmin( … MAX(0, -y w^Tx)) and not argmin( … MIN(0, -y w^Tx)) on p.28.
• Last year’s exam can be found here (PDF).
• The recitations on Tuesday, May 24th, are merged into one being held in NO C 60.
• The lectures on Tuesday, May 10th, and Wednesday, May 11th, have been cancelled.
• The lecture on Wednesday 4th of May has been cancelled.
• The rooms for the tutorials were changed/merged. For space reasons, we ask students who are scheduled on Fridays but need to go on Tuesdays (e.g. due to conflicts with other classes) to attend the Tuesday tutorial in NO C 60.
• Dummy project is available. For details check your email.
• The video recordings of the first week’s lectures are now available at ETH Videoportal.
• The files are password protected. To obtain the password you need to be inside the ETH network and click here. To establish a VPN connection click here.
• First class on 23.2. First tutorial/recitation on 1.3.

---

---

Matlab Demos

• Linear regression with gradient descent
• Linear regression for polynomials with gradient descent
• Gradient descent on multi-modal function with bold driver learning rate
• Kernelized linear regression demo
• Kernelized linear regression: Parameter effects
• Train Perceptron via SGD
• Train SVM via SGD
• L1-SVM
• Kernel Perceptron
• K-Nearest Neighbor
• Cost-sensitive Classification
• Multiclass: One-vs-All
• Multiclass: One-vs-One
• Training neural nets with backpropagation
• Universal Approximator Illustration
• Non-linear regression with neural nets and kernels
• Handwritten digit recognition with neural nets
• MNIST Data Set
• PCA

---

Java Applets

• Linear regression for polynomials
• Gaussian Mixtures

---

Video Lectures

• The video recordings are available at ETH Videoportal.

---

---

Project

Part of the coursework will be a project, carried out in groups of up to 3 students. The goal of this project is to get hands-on experience in machine learning tasks. The project grade will constitute 30% of the total grade. More details on the project will be given in the tutorials.

• Project information

---

Exam

The final exam takes place on August 13, from 9:00-11:00 in HIL F15 (Hönggerberg). The mode of examination is written, 120 minutes length. The language of examination is English. As written aids, you can bring two A4 pages (i.e. one A4 sheet of paper), either handwritten or 11 point minimum font size. The written exam will constitute 70% of the total grade.

---

Text Books

• K. Murphy. Machine Learning: a Probabilistic Perspective. MIT Press 2012
• C. Bishop. Pattern Recognition and Machine Learning. Springer, 2007 (optional)
• T. Hastie, R. Tibshirani, and J. Friedman. The Elements of Statistical Learning: Data Mining, Inference and Prediction. Springer, 2001. Available online
• L. Wasserman. All of Statistics: A Concise Course in Statistical Inference. Springer, 2004.

---

Matlab

• The official Matlab documentation is available online at the Mathworks website.
• If you have trouble accessing Matlab’s built-in help function, you can use the online function reference on that page or use the command-line version (type help at the prompt).
• There are several primers and tutorials on the web, a later edition of this one became the book Matlab Primer by T. Davis and K. Sigmon, CRC Press, 2005.

---