Kalender

Exportera Om schema på KTH

Logga in till din kurswebb

Du är inte inloggad på KTH så innehållet är inte anpassat efter dina val.

Ändra tidsperiod eller vy

Vecka 36 2015			Visa i Mitt schema

	Tis 1 sep 10:00-12:00	Lecture 1, Introduction
	HT 2015 Föreläsning Lärare: Giampiero Salvi, Atsuto Maki, Örjan Ekeberg Plats: F1	Structure of the Course What will the course cover? How are labs and examination handled? Learning Machines What do we mean by a "Learning Machine"? Supervised vs Unsupervised learning? What can learning algorithms be used for? How can a simple learning program be constructed? What is a Nearest neighbour classifier? Slides for Lecture 1 Slides for Lecture 1 (Part II)
	Tors 3 sep 17:00-19:00	Lecture 2, Decision Trees
	HT 2015 Föreläsning Lärare: Atsuto Maki Plats: E1	Anmärkning: Flyttad från kl 13-15 pga för liten sal Topics: What is a Decision Tree? When are decision trees useful? How can one select what questions to ask? What do we mean by Entropy for a data set? What do we mean by the Information Gain of a question? What is the problem of overfitting? Minimizing training error? What extensions will be possible for improvement? Slides for Lecture 2 Related reading: Chapter 8.1 from An Introduction to Statistical Learning (Springer, 2013) Gareth James, Daniela Witten, Trevor Hastie and Robert Tibshirani Available online: http://www-bcf.usc.edu/~gareth/ISL/
Vecka 37 2015			Visa i Mitt schema

	Tis 8 sep 10:00-12:00	Lecture 3, Challenges to machine learning
	HT 2015 Föreläsning Lärare: Atsuto Maki Plats: D2	Topics: Challenges to machine learning Model complexity and overfitting The curse of dimensionality Concepts of prediction errors The bias-variance trade-off Slides for Lecture 3 Related reading: Chapter 2 and 6.4 from An Introduction to Statistical Learning (Springer, 2013) Gareth James, Daniela Witten, Trevor Hastie and Robert Tibshirani Available online: http://www-bcf.usc.edu/~gareth/ISL/
	Tors 10 sep 13:00-15:00	Lecture 4, Regression Introduction
	HT 2015 Föreläsning Lärare: Atsuto Maki Plats: M2	Topics: Function approximation Linear regression RANSAC Nearest Neighbours regression Linear regression + regularization Ridge regression Lasso Slides for Lecture 4 Related reading: Chapter 3.1, 3.2, 3.5 and 6.2 from An Introduction to Statistical Learning (Springer, 2013) Gareth James, Daniela Witten, Trevor Hastie and Robert Tibshirani Available online: http://www-bcf.usc.edu/~gareth/ISL/
Vecka 38 2015			Visa i Mitt schema

	Tis 15 sep 10:00-12:00	Lecture 5, Probability I
	HT 2015 Föreläsning Lärare: Giampiero Salvi Plats: FR4	Topics: concepts of probability theory random variables and distributions conditional probabilities and Bayes rule Bayesian inference Slides for Lecture 5 Related reading: Prince, S.J.D., Part I (Chapters 2, 3, 5) If you want to know more a great book on these topics is Bishop, C. M. Pattern Recognition and Machine Learning, Springer.
	Tors 17 sep 13:00-15:00	Lecture 6, Probability II
	HT 2015 Föreläsning Lärare: Giampiero Salvi Plats: F2	Topics: estimation theory Maximum Likelihood, Maximum a Posteriori, Bayesian estimation Unsupervised learning and K-means Mixture of distributions and Expectation Maximization algorithm Slides for Lecture 06 Recommended reading: Prince, Chapter 3, 4, 7.4. (Book available in full PDF here)
Vecka 39 2015			Visa i Mitt schema

	Tis 22 sep 10:00-12:00	Lecture 7, Classification Introduction
	HT 2015 Föreläsning Lärare: Atsuto Maki Plats: K1	Topics: Naive Bayes Logistic regression Inference and decision Discriminative function Discriminative vs Generative model Slides for Lecture 7 Slides for Lecture 7 (part II) Related reading: Chapter 4 from An Introduction to Statistical Learning (Springer, 2013) Gareth James, Daniela Witten, Trevor Hastie and Robert Tibshirani Available online: http://www-bcf.usc.edu/~gareth/ISL/
	Tors 24 sep 13:00-15:00	Lecture 8, Classification with Separating Hyperplanes
	HT 2015 Föreläsning Lärare: Örjan Ekeberg Plats: FR4	Topics: Linear separation in high dimensional spaces Structural risk minimization Support Vector Machines Kernels for separating in a higher dimensional space Non-separable classes Slides from lecture 8 2 kommentarer
Vecka 40 2015			Visa i Mitt schema

	Tis 29 sep 10:00-12:00	Lecture 9, Artificial Neural Networks
	HT 2015 Föreläsning Lärare: Örjan Ekeberg Plats: FR4	Topics: Feed forward networks Using multiple processing layers Learning with Backprop Deep networks Slides from lecture 9
	Tors 1 okt 13:00-15:00	Lecture 10, Ensemble Methods
	HT 2015 Föreläsning Lärare: Atsuto Maki Plats: M1	Topics: Why combine classifiers? Bagging Decision Forests Boosting Slides for Lecture 10 Slides for Lecture 10 (full size) Related reading: Chapter 8.2 from An Introduction to Statistical Learning (Springer, 2013) Gareth James, Daniela Witten, Trevor Hastie and Robert Tibshirani Available online: http://www-bcf.usc.edu/~gareth/ISL/
Vecka 41 2015			Visa i Mitt schema

	Tis 6 okt 10:00-12:00	Lecture 11, Dimensionality Reduction
	HT 2015 Föreläsning Lärare: Atsuto Maki Plats: M2	We are going to revisit/resume Boosting at the beginning of the lecture. Topics: Unsupervised Learning Principal Component Analysis (PCA) Concept of subspace Similarity measures Subspace methods Slides for Lecture 11 Related reading: Chapter 10.2 from An Introduction to Statistical Learning (Springer, 2013) Gareth James, Daniela Witten, Trevor Hastie and Robert Tibshirani Available online: http://www-bcf.usc.edu/~gareth/ISL/
	Tors 8 okt 13:00-15:00	Lecture 12, Summary
	HT 2015 Föreläsning Lärare: Giampiero Salvi, Atsuto Maki, Örjan Ekeberg Plats: FR4	Slides for Lecture 12 Note: The scope of the exam is what has been covered in Lecture 1-11. 3 kommentarer

Kalender

Logga in till din kurswebb

Tidsperiod

Vy

Kontakt