- Home
- About Us
- People
- Research
- Services
- Best Practices Integrative Informatics Consultation Service (BPIC)
- High Performance Computing Facility (HPCF)
- Sequencing Informatics
- Microarray Informatics
- Proteomics Informatics
- Genetics-Genomics Informatics
- Cancer Informatics
- Centers of Excellence Informatics
- Multi-Modal & Integrative Studies
- Data Management and Mining
- Education
- Publications
- Industry Relations
Seminars and Tutorials
1. Tutorial in Support Vector Machines: Theory and Applications in Biomedicine.
Authors: Alexander Statnikov, Douglas Hardin, Isabelle Guyon, Constantin Aliferis. Materials about SVM Clustering were contributed by Nikita Lytkin.
Abstract:
This half-day tutorial will introduce support vector machines (SVMs) and their applications in biomedicine. SVMs are among the most important recent developments of machine learning and pattern recognition and have extensive applications in biomedicine and other fields. Unlike other approaches, these techniques are robust in data analysis with high variable-to-sample ratios and large number of irrelevant variables, they can learn efficiently very complex functions, and they employ powerful regularization principles to avoid overfitting.
A common obstacle in understanding and using SVMs is that they are mathematically challenging especially for biomedical researchers lacking extensive technical backgrounds. The tutorial is designed to enable all interested researchers grasp SVM fundamentals regardless of prior mathematical training. First, we will introduce basic principles behind SVMs in an intuitive manner. Then we will describe SVM-based algorithms for classification, regression, clustering and novelty detection, variable selection and dimensionality reduction. These algorithms are widely used and/or gaining popularity in biomedical applications. Throughout the tutorial we will provide case studies for each class of methods and give pointers to software implementations and additional literature. The knowledge gained in this tutorial will allow researchers to break the barriers of classical statistics and older pattern recognition and be able to conduct complex and high-dimensional analyses easily and efficiently.
Slides: Download pdf
Video (password-protected):
Lecture 1
Lecture 2
Lecture 3
Lecture 4
Lecture 5
Lecture 6
Copyright notice: Copyright (c) of its authors. No reproduction is allowed without written permission.
Notice about upcoming books: We are writing two books based on the materials of this tutorial: "A Gentle Introduction to Support Vector Machines in Biomedicine, Volume 1: Methods and Theory " and "A Gentle Introduction to Support Vector Machines in Biomedicine, Volume 2: Case Studies". The books will be published in 2010 by the World Scientific Publishing Co.
History and availability: This tutorial was first presented in the AMIA Fall Conference 2008. It will be presented again in the AMIA Fall Conference 2009. In addition a video recording will be made available through this page in December 2009.
2. Special Faculty/Staff Tutorial on Local causal Discovery and Feature selection methods (Spring 2009)
Author: Constantin Aliferis
Abstract: This tutorial introduced to CHIBI faculty and staff, some fundamental algorithms and principles for causal graph-based local causal discovery and feature selection. The tutorial may be continued/expanded and opened to the general public given enough interest. Please contact Constantin Aleferis or Alexander Statnikov
3. Tutorial in Machine Learning for Biomedical Decision Support and Discovery
Authors: Constantin F. Aliferis and Ioannis Tsamardinos. Alexander Statnikov contributed supplementary information.
Abstract: The purpose of this tutorial is to (i) help participants develop a solid understanding of some of the most useful machine learning methods, (ii) give several examples of how these methods can be applied in practice, and (iii) provide resources for expanding the knowledge gained in the tutorial.
Slides:
- Tutorial overview and goals
- Importance of Machine Learning for discovery and decision support system construction
- A framework for inductive Machine Learning
- Generalization and overfitting
- Quick review of data preparation and model evaluation
- Bayesian classifiers
Presentation file II: Neural Networks
Presentation file III: Case Study on Predicting Breast Cancer Invasion with Artificial Neural Networks on the Basis of Mammographic Features
Presentation file IV: Support Vector Machines
Presentation file V: Quick Review of Additional families (K-Nearest Neighbors, Clustering, Decision Tree Induction, Genetic Algorithms)
Presentation file VI: Causal Discovery Methods Using Causal Probabilistic Networks
Presentation file VII: Feature selection
Presentation file VIII: Case studies
- Conclusions and wrap-up
- Resources for machine learning
- Questions & feedback
Copyright notice: Copyright (c) of its authors. No reproduction is allowed without written permission.
History and availability: This tutorial was first presented in the MEDINFO Conference 2004 and prior to this in AMIA Fall Symposium 2003.
4. Upcoming tutorials (to be released in 2010):
Next generation sequencing informatics
This tutorial will introduce fundamental concepts for the generation and analysis of Next-Generation Sequencing data from popular instruments such as the Illumina GA and the Roche 454. The tutorial will cover the fundamentals of the employed sequencing technology, informatics protocols for analysis of data, and examples from a spectrum of applications including, Chip-seq, epigenetics, microbiomics, RNA-seq and other types of studies.
Statistical genetics
This tutorial will introduce the basic research designs, study types and analytics for high-throughput genetic studies.
For recommendations for new tutorials please contact Yuval Kluger: Yuval.Kluger@med.nyu.edu
