ROE 501--- Applied Machine Learning

Izmir Katip Çelebi University, Graduate School of Natural Sciences

Spring 2020

Syllabus

Course Code and Name

ROE 501- Applied Machine Learning

Instructor

Associate Professor Aytuğ ONAN

E-mail

aytug.onan@cbu.edu.tr

aytugonan@gmail.com

Course Day and Time

 Tuesday, 15.00-17.00

Course Website

http://aytugonan.cbu.edu.tr/ROE501_index.html

Objectives

In the modern IT world, businesses often have access to large amounts of data collected from customer management systems, web services, customer interaction, etc. The data in itself does not bring value to the business; we must bring meaning to the data to create value. Data mining and machine learning is an area within computer science with the goal of bringing meaning to and learning from data. This course will focus on applied machine learning, where we learn what algorithms and approaches to apply on different types of data.

Tentative Course Outline

 

Week#1: Course Introduction

Week#2: Introduction to Machine Learning

Week#3: Loss Functions

Week#4: Neural Networks

Week#5: Neural Language Models (Project Proposal!!)

Week#6: Recurrent Neural Networks

Week#7: No Class (Midterm Week!!)

Week#8: LSTM

Week#9: Sequence to Sequence Models

Week#10: Convolutional Neural Networks

Week#11: Domain Adaptation

Week#12: Student Presentations

Week#13: Student Presentations

Week#14: Student Presentations

 

Textbook

Mining of Massive Datasets
Leskovec, Jure & Rajaraman, Anand & Ullman, Jeffrey David (2014), Cambridge University Press, 476 pages.
Available for free online here.

Deep Learning
Goodfellow, Ian & Bengio, Yoshua & Courville, Aaron (2016), MIT Press, 781 pages.
Available for free online here.

Supplementary Materials

Kuncheva, L. I. (2004). Combining pattern classifiers: methods and algorithms. John Wiley & Sons.

Evaluation

Research Project:

Research Presentation: 30%      (Week#11-14)

Final Report: 70%                       (Week#11)

 

This will be a  project involving choosing an interesting machine learning question, finding relevant data, using an appropriate toolbox (Matlab, TensorFlow, R, Scikit, Keras, etc.) to answer the question, writing it up, and presenting it to the class. This should be done as an individual.