CSC 4510/9010: Applied Machine Learning
Mendel G88, Fall, 2016
Dr. Paula Matuszek
Adjunct Professor, Villanova
Office:  Mendel 288.
E-mail: Paula.Matuszek@villanova.edu or Paula.Matuszek@gmail.com
Phone: (610) 647-9789
TA: Shiva Beesu.  Office Hours Weds, Thurs 10-1.

Schedule:

Machine learning is a fast-moving field, and this is a new class arrangement, so this schedule is a work in progress. The midterm and final dates are definite.  Everything else is subject to change as I see how the semester goes.  Links to presentations and assignment announcements will be posted as we reach them.

Date(s) Topic Presentations Reading Assignment
Aug 25: Intro, architecture of a learning system. Idea of a model. Brief intro to Weka. ML Introduction
Weka Introduction
W,F&H:  pp1-21
Assignment 1:  Weka Installation.   Due Aug 31.
Aug 30: Weka.  Supervised and unsupervised learning Supervised and Unsupervised Learning
W, F&H: pp21-28
none
Sept 1: Supervised Learning, inputs, kinds, good training examples.  Piazza. Kinds of Machine Learning.  Inputs to ML systems.
W, F&H, pp-28-50.  This is through section 2.3.
Assignment 2:  Running a Decision Tree.  Due Sept 7.
Sept 6, 8: Decision trees and J48. Decision Trees.  Weka and ARFF . Restaurant1.
Restaurant2.
W, F&H, sections 4.3 and 6.1.
none
Sept 13: Evaluation of ML Results Evaluation.  
W, F&H, Sections 5.1, 5.3, 5.7.
Assignment 3:  An Interesting Question
Sept 15 K Nearest Neighbor Classifiers KNN  Using Weka Models
W, F&H, Sections 4.7, 6.5.
Assignent 4:  KNN and Applying Weka.  irissubset.arff
Sept  20, 22: Neural Nets, Weka Perceptrons, Deep Learnng Neural Nets.    Deep LearningNeural Nets Lab.  Interesting Questions.
W, F&H, pp 236-241.  Multilayer Perceptrons in section 6.4.
Assignment 5:  NN Lab and Interesting Questions.
Sept 27: Rules, Visualizing Classifier Boundaries RulesVisualization. W,F&H:  sections 3.4, 4.1, 4.4, 17.3. National Voter Registration Day. 
Sept 29, Oct 4: Regression-based ML. Weka Logistic and Linear Regression Linear Regression Contact info for projects.  Logistic Regression.    irispetal.arff W,F&H:  sections 4.6, 5.8
midterm prep!
Oct 6: Midterm Pre-post notesMidterm Notes.


Oct 11-13: fall break


Oct 18-20 SVMs. Weka SMO SVM PresentationSVM LabText Lab. CrudeTrain.arff
W, F&H pp 223-228, 462-467, 578-582
Assignment 6.
Oct 25 Naive Bayes NB Presentation.
Sections 4.2, 6.7

Oct 27, Nov 1, 3: Unsupervised learning. Clustering. Weka Clustering   Clustering 1 presentationProject Information. Interpreting clusters presentationClustering 2 presentation. ClusterTest.arff. bmw-browsers.arff.
Sections 4.8, 6.8
Assignment 7.
Assignment 8.
Nov 8: Unsupervised learning. Dimensionality Reduction.. Dimensionality Reduction presentationFinding Data Sources.

Vote!
Nov 10: Putting It All Together Putting It Together Exercise
Assignment 9.
Nov 14.: Semi-Supervised Learning Semi-supervised Learning Presentation


Nov 17: Reinforcement Learning Reinforcement Learning Presentation

Presentation Information
Nov 22: Transfer Learning Transfer Learning Presentation
Survey on Transfer Learning
Nov 24:


Nov 29: Convolutional Networks, AdaBoost AdaBoost PresentationConvolutional Neural Networks.
A Brief Introduction to Boosting

Dec 1: Inductive Learning Inductive Learning

Prepare for Presentation!
Dec 6: Student Presentations.


Dec 8: Last class.  Student evaluations.  Student Presentaion.  Other machine learning tools. Some current directions. SummaryNotes on Final.


Dec 16: Final, 2:30-5:00