<html>
<head>
<meta http-equiv="content-type" content="text/html; charset=ISO-8859-1">
</head>
<body bgcolor="#FFFFFF" text="#000000">
Dear colleagues,<br>
<div class="moz-forward-container">
<div class="moz-forward-container"> <br>
The Technical University of Madrid (UPM) will once more organize
the summer<br>
school on 'Advanced Statistics and Data Mining' in Madrid
between June<br>
24th and July 5th. This year's programme comprises 12 courses
divided<br>
into 2 weeks. Attendees may register in each course
independently.<br>
<br>
Early registration is now *OPEN*. Extended information on course
programmes, <br>
price, venue, accommodation and transport is available at the
school's website:<a moz-do-not-send="true"
href="http://www.dia.fi.upm.es/ASDM"><br>
<br>
http://www.dia.fi.upm.es/ASDM</a><br>
<br>
Please, send this information to your colleagues, students, and
whoever may <br>
find it interesting. <br>
<br>
Best regards,<br>
<br>
Pedro Larrañaga, Concha Bielza and Pedro L. López-Cruz.<br>
-- The coordinators of the school.<br>
<br>
<br>
*** List of courses and brief description ***<br>
<br>
* Week 1 (June 24th - June 28th, 2013) *<br>
<br>
1st session: 9:30 - 12:30<br>
Course 1: Bayesian networks (15 h)<br>
Basics of Bayesian networks. Inference in Bayesian
networks.<br>
Learning Bayesian networks from data. Real applications.<br>
<br>
Course 2: Statistical inference (15 h)<br>
Introduction. Some basic statistical test. Multiple
testing.<br>
Introduction to bootstrap methods. Introduction to Robust
Statistics.<br>
<br>
2nd session: 13:30 - 16:30<br>
Course 3: Supervised pattern recognition (15 h)<br>
Introduction. Assessing the performance of supervised
classification <br>
algorithms. Preprocessing. Classification techniques.
Combining<br>
multiple classifiers. Comparing supervised classification
algorithms.<br>
<br>
Course 4: Multivariate data analysis (15 h)<br>
Introduction. Data examination. Principal component
analysis.<br>
Factor Analysis. Multidimensional scaling. Correspondence
<br>
analysis. Tensor analysis. Multivariate Analysis of
Variance. <br>
Canonical Correlation Analysis. Latent Class Analysis.<br>
<br>
3rd session: 17:00 - 20:00<br>
Course 5: Neural networks (15 h)<br>
Introduction. Perceptrons. Training algorithms.
Accelerating<br>
convergence. Useful tricks for MLPs. Deep networks.
Practical<br>
data modelling with neural networks.<br>
<br>
Course 6: Feature Subset Selection (15 h)<br>
Introduction. Filter approaches. Wrapper methods. <br>
Embedded methods. Drawbacks and future strands.<br>
Practical session.<br>
<br>
* Week 2 (July 1st - July 5th, 2013) *<br>
<br>
1st session: 9:30 - 12:30<br>
Course 7: Time series analysis (15 h)<br>
Introduction. Probability models to time series.
Regression and<br>
Fourier analysis. Forecasting and Data mining.<br>
<br>
Course 8: Hidden Markov Models (15 h)<br>
Introduction. Discrete Hidden Markov Models. Basic
algorithms<br>
for Hidden Markov Models. Semicontinuous Hidden Markov
Models.<br>
Continuous Hidden Markov Models. Unit selection and
clustering.<br>
Speaker and Environment Adaptation for HMMs.<br>
Other applications of HMMs.<br>
<br>
2nd session: 13:30 - 16:30<br>
Course 9: Bayesian classifiers (15 h)<br>
Discrete predictors. Gaussian Bayesian networks-based
classifiers.<br>
Other Bayesian classifiers. Bayesian classifiers for:
positive and<br>
unlabeled data, semi-supervised learning, data streams,
temporal<br>
data.<br>
<br>
Course 10: Unsupervised pattern recognition (15 h)<br>
Introduction. Prototype-based clustering. Density-based<br>
clustering. Graph-based clustering. Cluster evaluation.<br>
Miscellanea.<br>
<br>
3rd session: 17:00 - 20:00<br>
Course 11: Support vector machines, regularization and convex
optimization (15 h)<br>
Introduction. SVM models. SVM learning algorithms. Convex<br>
non differentiable optimization.<br>
<br>
Course 12: Hot topics in intelligent data analysis (15 h)<br>
Multi-label and multi-dimensional classification.
Multi-dimensional<br>
classification and multi-output regression. Advanced
Clustering. <br>
Partially supervised classification with uncertain class
labels. <br>
Directional statistics. Spatial point processes.<br>
<br>
</div>
</div>
</body>
</html>