(2nd Semester Course of the MSc Data Science and Machine Learning from the NTUA)
• Vasilis Maglaris (http://www.netmode.ntua.gr/profiles/bmaglaris_en.html)
Overview of Optimization Algorithms in Machine Learning: Supervised, Unsupervised, Competitive, Enhanced Learning. Linear & Logistic Regression, parameterization of Neural Networks with supervised learning, Back-Propagation Algorithm
Reducing Complexity through Unsupervised Learning: K-Means Clustering, Principal Components Analysis (PCA)
Basic Concepts of Statistical Engineering in Machine Learning: Markov chains, situation classification, transition probabilities, Chapman-Kolmogorov equations, reproducibility - transience, unchanged distributions, asymptotic behavior. Monte Carlo Markov Chain Simulation Methods, Metropolis - Hastings Algorithm. Simulated Annealing, Gibbs sampling. Generative Models, Boltzmann Machine, Restricted Boltzmann Machine (RBM), Deep Belief Nets (DBN)
Reinforcement Learning and Dynamic Programming: Markov Decision Processes, Bellman’s Optimality Criterion, Value and Policy Iteration algorithms. Dynamic programming approximation methods, Q-Learning
Routing Learning for Internet Routing: Bellman - Ford Algorithm, Border Gateway Protocols (BGP)
Probabilistic Classification: Bayes Rule, Approximate Methods - Naive Bayes Algorithm
Core Algorithms and Pattern Resolution: Cover Theorem, Applications in Radial-Basis Function (RBF) Networks, Hybrid Learning, Support Vector Machines (SVM)
Decision Trees: CART (Classification And Regression Trees), Gini Index, Random Forests, Bagging (Bootstrap & aggregating) Algorithms
Algorithms based on Sequential Related Data Learning Models: Time-series & Speech Processing Datasets, Recurrent Neural Nets (RNN), Long-Short Term Memory Networks (LSTM)
- Simon Haykin, “Neural Networks and Learning Machines”, Third Edition, Pearson Education, 2009
- Simon Haykin, “Νευρωνικά Δίκτυα και Μηχανική Μάθηση”, Τρίτη Έκδοση, Παπασωτηρίου, 2010 (Ελληνική μετάφραση)
- Μιχάλης Λουλάκης, “Στοχαστικές Διαδικασίες”, ΣΕΑΒ 2015
- Βασίλης Μάγκλαρης, “Σημειώσεις Μαθήματος Συστήματα Αναμονής”, Συλλογή διαφανειών για το προπτυχιακό μάθημα της ΣΗΜΜΥ – ΕΜΠ, 2018 http://www.netmode.ntua.gr/courses/undergraduate/queues/documents/Queuing_Systems_2018.pdf
- Kevin P. Murphy, “Machine Learning: A Probabilistic Perspective”, MIT Press, 2012
- Ian Goodfellow, Yoshua Bengio and Aaron Courville, “Deep Learning”, MIT Press, 2016
- Daniel Jurafsky and James H. Martin, “Speech and Language Processing: An Introduction to Natural Language Processing, Computational Linguistics, and Speech Recognition”, Third Edition draft, 2018
- Andrew Ng, "CS229 Lecture Notes", Stanford University, Fall 2018 http://cs229.stanford.edu/notes/cs229-notes1 -James Gareth, Daniela Witten, Trevor Hastie and Robert Tibshirani, "An Introduction to Statistical Learning with Applications in R", Springer 2013, https://www-bcf.usc.edu~gareth/ISL/ISLR%20First%20Printing.pdf
- Richard Sutton and Andrew Barto, "Reinforcement Learning: An Introduction", Second Edition, MIT Press, 2018
- Christopher Bishop, “Pattern Recognition and Machine Learning”, Springer 2006
- Tom Mitchell, “Machine Learning”, McGraw Hill 1997 http://www.cs.cmu.edu/~tom/mlbook.html
- Frank Kelly, "Reversibility and Stochastic Networks", Wiley, 1979 http://www.statslab.cam.ac.uk/~frank/BOOKS/book/whole.pdf
- Sheldon Ross, "Applied Probability Models with Optimization Applications", Dover, 1992
This is the repository for the postgraduate course Stochastic Processes & Optimization in Machine Learning. This course is included in the Data Science & Machine Learning (DSML) program of the National Technical University of Athens (NTUA).
Our 2021 course will include the following exercises provided as Jupyter Notebooks:
- Linear Regression and Polynomial Regression
- Logistic Regression, K-means Clustering and Principal Component Analysis (PCA)
- Markov Chains and Simulation (heavily based on the Stochastic Processes course of the 6th semester in ECE NTUA)
- The Metropolis-Hastings Algorithm
- Simulated Annealing and Restricted Boltzmann Machine (RBM)
- Markov Decision Processes and Q-Learning
- Bellman-Ford Algorithm (Application in the BGP protocol)
- Naive Bayes Classifier (Application in DNS DDoS attack mitigation: the DNS Water Torture Attack use case)
- Radial Basis Function (RBF) and Support Vector Machine (SVM)
- Decision Trees and Random Forests
- Long Short-Term Memory (LSTM), Application in the DNS DDoS attack mitigation: the DNS Water Torture Attack use case
Note: Some exercises are taken from online sources.