Lecturer: Riccardo Bellazzi   Silvana Quaglini  

Course name: Data Mining And Biomedical Decision Support System
Course code: 503001
Degree course: Bioingegneria
Disciplinary field of science: ING-INF/06
L'insegnamento è caratterizzante per: Bioingegneria
University credits: ECTS 12
Course website: n.d.

Specific course objectives

The course aims to provide students with methodological skills and techniques to: * use in biomedical applications a large class of algorithms that are able to learn decision rules from data and automatically improve their performance based on experience. * Modeling of complex medical problems, in which you require to make decisions under uncertainty and / or taking into account the preferences of the patient. We will discuss diagnostic problems, therapeutic and monitoring ones. The student at the end of the course, should be able to: * soundly apply data mining approaches to learn decision rules from data * use machine learning software tools and statistical packages * formalize a decision problem, identifying the variables in the domain and choosing the formalism more suitable, both for the acquisition of knowledge (interaction with the medical counterpart for the construction of the model and interaction with the patient for the elicitation of preferences), either for the purpose of solving the problem. Among the classes of decision problems, particular emphasis will be given to economic evaluations prior to decision goodwill or less of a health program. The course will include both lectures and practical hands-on computer lessons.

Course programme

Learning decision rules - supervised learning

• Introduction: Machine Learning and Data Mining in the biomedical sciences.
• Areas of application of automatic methods for classification: diagnosis, prognosis, research
• The basic concepts: examples, instances, attributes, and representation of decision rules
• Decision Trees: learning techniques for pruning
• Bayesian methods: Naive Bayes discriminant analysis
• Regression models: linear model, logistic regression, neural networks, support-vector machines
• Method and k-nearest distance measures
• Learning of rules covering methods, beam-search methods
• Techniques of feature selection. Information gain and Relief
• Evaluation of learning algorithms and problems of evaluation in the biomedical field
• Training and Testing. Accuracy, calibration, sensitivity and specificity, precision and recall, F measure
• Methods for performance evaluation. Cross Validation, Bootstrap and ROC curves.

Unsupervised learning

• Association Rules
• Clustering methods: K-means, K-medoids, hierarchical clustering, self-organizing maps
• Evaluation of the results of the clustering methods
• Applications of data mining in bio-medicine: diagnosis, prognosis, classification, functional genomics

Practical Activities

• The CRISP methodology for data mining in bio-medicine.
• Hands-on with computer programs: Orange, Weka and Matlab for the solution of classification problems.

Introduction to decision-making
The uncertainty and preferences as fundamental problems of medical decision-making and brief review of the basic concepts of probability theory

Decision Theory

• Quantifying the value of an outcome (health status, life expectancy)
• Methods for the quantification of utility (rating scale, standard gamble, time-trade-off
• Expected utility of a decision
• the QALY (Quality-adjusted life-years)
• probabilistic dominance
• decision Trees
• methodologies for the construction and resolution of decision trees;
• use of a software for the management of decision trees;
• representation of Markov processes within a decision tree;
• sensitivity analysis and threshold, univariate and multivariate analysis;
• Influence diagrams
• Analogy with probabilistic networks
• Methodologies for the construction and resolution;
• Use of a software for building and solving influence diagrams
• Economic evaluations of health programs
• Cost-effectiveness analysis, cost-benefit, cost-utility;
• the points of view of society ', the NHS, the third-party payer
• critical reading of a literature article

Course entry requirements

Basic knowledge of statistics and probability theory. Basic knowledge of informatics and statistical software tools

Course structure and teaching

Lectures (hours/year in lecture theatre): 90
Practical class (hours/year in lecture theatre): 0
Practicals / Workshops (hours/year in lecture theatre): 0

Suggested reading materials

T. Mitchell. Machine Learning. Mc Graw Hill.

P. Tan, M. Steinbach, V. Kumar. Introduction to data mining. Addison Wesley.

I. Witten, E. Frank. Data mining. Morgan Kaufmann.

M.C. Weinstein, H.V. Fineberg. Analisi decisionale. L. Franco Angeli Editore.

David M. Eddy. Clinical Decision Making. Jones and Bartlett Publishers.

R. Bellazzi. Slides.

S. Quaglini. Slides.

Testing and exams

Learning decision rules: A written exam and an essay about a data mining problem to be carried on with machine learning software on a data set provided to the students. Decision analysis: Written test and oral exam