CS6510: Applied
Machine Learning
Fall 2016
Instructor: Vineeth N
Balasubramanian
Class Schedule: Sat 9:00 am –
12:30 pm
Class Location: LH-1
TAs:
Arghya Pal, Supriya Pandhre, Mausam Jain, Akilesh B, Hrishikesh Vaidya, Sahil
Manocha
OBJECTIVE: This course
will introduce core statistical machine learning methods with a relatively
low-math approach, emphasizing applied problem-solving and engineering. At the
end of the course, the student will have a sound understanding of various
machine learning methods for prediction tasks, their intuition, and how they
can be applied to real-world problems.
TOPICS (Tentative): Classification (Naive Bayes, k-NN, SVM, Neural
Networks, Decision Trees, Logistic Regression, Ensemble Methods), Regression
(Linear, Non-linear, k-NN, SVR), Clustering (k-means, DBSCAN, hierarchical),
Dimensionality Reduction (PCA, MDS, Isomap), Gaussian Mixture Models, EM,
Feature Selection, Model Selection and
Performance Evaluation (Cross-Validation, Bootstrap, ROC), Time series analysis
methods
The course will
also have a significant hands-on component involving in-class demonstrations,
practice exercises and Kaggle-style class contests (and potentially an ML
hackathon conducted by Microsoft).
ELIGIBILITY: This
course is intended for EMDS 2nd year students, and BTech 3rd
year students (CSE/ES)
PRE-REQUISITES:
Probability, Linear Algebra, Calculus
LECTURES: Lec1 – Intro (PDF)
REFERENCES:
1. Witten,
I. H., Frank, E., and Hall, M. (2011). Data Mining: Practical Machine
Learning Tools and Techniques, third edition, Elsevier: San Francisco, ISBN
978-0-12-374856- 0 (The book behind Weka, the popular
ML toolkit)
2.
Pattern
Recognition and Machine Learning by C. Bishop
OTHER USEFUL
RESOURCES:
-
Elements of Statistical
Learning by Hastie, Tibshirani and Friedman
- The lecture notes from the old Learning from
Data course are useful, although they contain more mathematical detail than we
will do in this course.
- A Few Useful Things to Know about Machine Learning by
P. Domingos
- Coursera course on Machine Learning by Andrew Ng