Between 2006 and 2010, I was privileged enough to be in charge of giving a number of lectures on the topics of my expertise, among others especially teaching the course “Data Mining Algorithms” at the postgraduate program “Applied Mathematical Sciences” of NTUA’s School of Applied Mathematics and Physics. The whole course was being performed through a blog. You may find out more about the process here, while more details on the lectures’ content is provided below.
Data Mining Algorithms
• Context: A course for the post-graduate program ‘Applied Mathematical Sciences’ of the School of Applied Mathematics and Physics, National Technical University of Athens
• Tasks assigned: Teaching Instructor
• Course contents: Attribute selection, filters, validation (holdout, stratification, cross-validation, bootstrap), classification rules, decision trees, naive Bayes, instance-based representation, association rules, neural networks, support vector machines, bagging, boosting, stacking, implementation in Weka
• Lecture notes (in greek):
– Lecture 01, Introduction (.pdf)
– Lecture 02, Data Components, Visualization & Exploration (.pdf)
– Lecture 03, Data Preprocess & Attribute Selection (.pdf)
– Lecture04, Knowledge Representation, Credibility & Evaluation (.pdf)
– Lecture05, Algorithms: Trees & Rules (.pdf)
– Lecture06, Algorithms: Bayes, Association Rules & Instance-based Clustering (.pdf)
– Lecture07, Algorithms: Support Vectors, Neural Nets, Meta-algorithms (.pdf)
Introduction to Forecasting
• Context: A couple of lectures taught at the undergraduate courses ‘Logistics’ and ‘Production Management I’, School of Mechanical Engineering, National Technical University of Athens
• Tasks assigned: Teaching Instructor
• Lectures contents: Definition and terminology, applications, implications in logistics chain and decision support, classification of available tehniques, judgmental and causal methods, quick introduction to econometrics (covariance and correlation coefficient, regression, R2 coefficient, implementation in MS Excel), quick introduction to time-series analysis (time-series modeling and decomposition -trend, seasonality and randomness-, smoothing and moving averages, Holt-Winters model, errors, covariance, autocorrelation and stationarity, implementation in MS Excel), introduction to neural networks, implementation in Matlab
• Lecture notes (in greek): Introductory Lecture to Forecasting (.pdf)
also: Extended Lecture Notes on Prediction Markets (.pdf)
Introduction to Data Mining: From Data to Knowledge
• Context: A lecture taught at the undergraduate course ‘Management Information Systems’, School of Mechanical Engineering, National Technical University of Athens and at the same titled course of ‘Athens MBA’, a joint effort of National Technical University of Athens and Athens University of Economics and Business
• Tasks assigned: Teaching Instructor
• Lecture contents: knowledge management and the need for data mining in the modern enterprise, definition and the process from data to information to knowledge, patterns, case studies, concepts, instances and attributes, classification, association, clustering and numeric prediction, supervised and unsupervised algorithms, data transformations, trial & error, decision tables, trees and rules, association rules, linear regression, instance-based representation and clusters, training and testing, cross-validation, bagging, boosting and stacking, introduction to and demonstration of Weka
• Lecture notes (in greek): MIS, Intro to Data Mining (.pdf)