Özyeğin University, Çekmeköy Campus Nişantepe District, Orman Street, 34794 Çekmeköy - İSTANBUL

Phone : +90 (216) 564 90 00

Fax : +90 (216) 564 99 99

E-mail: info@ozyegin.edu.tr

M.Sc. in Data Science

M.Sc. in Data Science

Overview

Along with advancing technology, data storage and processing techniques, Data Science is one of the fastest developing areas of the 21st century and is critical to any organization or company that wants to achieve the ability to collect, understand and analyze appropriate data for maintaining its competitiveness. In this context, data science has taken the place among the professions of today and the future.

The aim of our Master of Science degree in Data Science is to train “Data Scientists” who have the latest knowledge, concepts, data processing and analysis techniques, and great team working skills. With these characteristics, our graduates can contribute to industry applications with the competencies they have gained, and they will be entrepreneurial pioneers of change and development by creating the qualified scientific manpower in both domestic and international business world.

Özyeğin University, with the visions of being an entrepreneurial research university and leading by developing the ecosystem in digital transformation; offers a powerful theoretical and practical education to its students from different disciplines. The world class education also supports the culture of self-learning by using new generation tools and online resources.

The program aims to build a strong international student body with both Turkish and international students; and will be further enriched with intensive international activities and collaborations to expose students to a wide array of opportunities for international and cross-cultural interactions.

Data Science graduate students, together with our other graduate students, will have the opportunity to convert their thesis subjects to new business areas or to use them for the benefit of the institution or company to which they are affiliated.

 

DATA SCIENCE REQUIRED COURSES

Students are required to take at least 3 courses from the pool below. If the students take more than 3 courses from this pool, they can count them as the elective courses within the program requirements.

 

DS 528 Customer Analytics

The aim of this course is to provide a background for students to use probabilistic, statistical and optimization-based models developed to model customer behavior. Throughout the course, students develop strong intuition about the importance of different models developed for various conditions, suitability and performance of the problem; and use the “timing”, “counting”, “selection” models to gain comprehensive knowledge of customer retention, customer churn, life expectancy, cross-selling, up-selling and next product to buy.

DS 529 Predictive Analytics

The aim of this course is to introduce the predictive analytical methods used in the prediction of various parameters in businesses such as customer demand and churn prediction. Quantitative prediction techniques, time series analysis, moving average, exponential smoothing methods, regression analysis (simple, multiple and logistic) and analysis of prediction errors are among the topics covered. Applications of these techniques are examined using real data, case analysis and statistical software.

DS 538 Mathematical and Statistical Foundations of Analytics                              

This course aims to teach the foundations of probability and statistics and methods of extracting information from data. The course focuses on data analysis, probability theory and statistical inference; After providing the basics of these topics, regression analysis is studied extensively. Special topics such as curve fitting and prediction are also covered throughout the course. Probability modeling and statistical learning are covered with appropriate use case scenarios. Python (or R software) is used for the application of theoretical knowledge.

DS 540 Machine Learning with Python

This course covers some basic models and algorithms for regression, classification and clustering. Topics include linear and logistic regression, regularization, kNN, LDA, QDA, probability (Bayesian) extraction, SVMs and kernel methods, clustering and feature selection. Using Python programming language, students handle machine learning problems.

DS 555 Data Science and Business Strategy

This course provides a general framework for the fundamental / destructive changes in business strategy caused by recent developments in data science. Topics include (i) data-driven / analytical thinking, (ii) business problems and related data science solutions, (iii) strategic thinking and data science / machine learning applications, (iv) extracting business views from models, (v) organizational transformation and challenges. A wide range of practical applications in several sectors will be discussed as case studies.

DS 581 Data Mining

The aim of this course is to equip students with the knowledge of basic data mining techniques. These techniques include descriptive methods such as clustering, basket analysis, row analysis, and predictive methods such as decision trees and logistic regression. Theoretical courses are supported by practical studies and provide basic skills in using a data mining package program. At the end of the course, students will be able to apply data mining methods to real-life problems and be able to use some basic methods for problem solving.

DS 587 Data Science

The aim of this course is to enable students to learn about data science and analytical data analysis, data conversion, data engineering technologies (MapReduce, Hadoop, Spark), etc. With this knowledge, students will be able to solve real business problems (such as predicting customer churn or response rates) using predictive algorithms. They also learn how to develop data-driven models, implement these models using one of the widely preferred software (R, Python, etc.) in the field of data science and share the results effectively.

DS 588 Data Science and Algorithms

This course focuses on data types and algorithms that form the foundations of data analytics. In order to do this, statistical and mathematical foundations of algorithms are explained. Today, due to the variable nature of data, analyzing data is a complex process and requires programming and software knowledge. This course includes software such as Stata, Matlab and Python programming languages. On the algorithm side, data preparation and reduction techniques, data representation, basic data science and machine learning methods are explained. Real data is studied in the classroom practices.

 

DATA SCIENCE ELECTIVE COURSES

IE 501 Linear Programming and Extensions

This course aims to provide students with a sound theoretical background on linear optimization and its extensions in other optimization areas.  It builds upon previously acquired introductory knowledge on Linear Programming (LP), which includes development of LP models and the workings of the Simplex Algorithm.  Students develop a deeper understanding of the mathematical underpinnings of the Simplex Algorithm and linear optimization in general. Topics include optimality conditions, duality theory, and methods for large scale optimization. Students also practice with using CPLEX, a state-of-the-art optimization software, through a project of their choice.

IE 502 Integer Programming

The aim of this course is to introduce both the practice and the theory of integer programming. The course covers integer programming’s scope and applicability, MILP models, linear inequalities and polyhedral, split inequalities, intersection cuts, valid inequalities, lift-and-project procedure, Benders’ decomposition, and enumeration.

IE 503 Nonlinear Programming

The aim of this course is to model optimization problems in which the results of decisions are connected with nonlinear relationships and to teach related solution algorithms. Topics include convex analysis, necessary and sufficient conditions for optimality, analysis of equality and inequality constraints, confidence interval methods, Nelder-Mead method, Newton method and its derivatives, loss function methods, etc. In this course, theoretical analysis and practical applications will be equally weighted, and the students will be expected to apply mathematical proofs by using a general-purpose computer programming language or commercial optimization software.

IE 504 Heuristic Methods

A large number of real-life optimization problems in engineering are complex and difficult to solve with classical optimization techniques. They cannot be solved in an exact manner within a reasonable amount of time; therefore, using approximate algorithms is the main alternative to solve this class of problems. In this course, students will learn a variety of heuristic methods, which are used to find good but not necessarily optimal solutions to difficult optimization problems within a reasonably short time. The course aims to provide an understanding of the general characteristics and limitations of a wide range of classical and modern heuristics; introduce techniques to design and implement heuristics; and present applications of heuristic methods to different problems.

IE 512 Network Science

The course covers definitions of basic concepts in graph theory and networks. It covers mathematics of networks, measures and metrics required for network analysis and how social networks are formed. It involves communities over social networks and processes such as finding critical components, cascading effects, information diffusion, etc. Such problems are relevant in a variety of settings from marketing to telecommunications. 

IE 521 Operations Research and Applications

This course explores issues in the design, implementation, and evaluation of Intelligent User Interfaces that use artificial intelligence technologies such as machine learning, computer vision and pattern recognition. The purpose of the course is to cover three aspects of intelligent user interface design including the development process, supporting device technologies and the related algorithmic knowledge. The course will be organized around reading and discussion of seminal and recent papers from the related literature.

 IE 531   Probabilistic Analysis

The aim of this course is to teach the fundamentals of probability theory used in engineering and applied sciences. Probability theory topics include sample space events, probabilistic models, random variables, conditional probability, discrete and continuous distributions, expectations, joint / compound distributions, law of large numbers and central limit theorem. Topics include inferential statistics, sampling, point and confidence interval estimates, and hypothesis testing. The focus of the course will be on practical applications and the algorithms learned in the course will be expected to be implemented by using specialized statistical software.

IE 532 Stochastic Models

This course provides an introduction to stochastic modeling. It covers various topics including conditional probability and expectation, random walks, Poisson processes, renewal theory, discrete and continuous time Markov chains, and queuing theory and reliability. Basic concepts, rules and approaches taught in the course will be introduced through real-life applications in inventory control, production, finance, and communication systems.

IE 534 Advanced Statistics                                                                                                                      

The aim of the course is to teach the students the statistical models used in engineering and applied sciences, how to interpret these models and their results. The emphasis of the course is on the basic principles of statistical inference. The course material also includes basic regression analysis and linear models. Topics include the concept of statistical estimator, point estimation, confidence intervals, regression analysis, Cramer-Rao lower bound, hypothesis testing, Goodness of fit tests, variance analysis, experimental design methodology, classification theory and clustering analysis. Applications of the methods taught in the course in engineering and applied sciences will also be introduced.

IE 535 Advanced Simulation

The aim of this course is to teach the basics of simulation methodology and to show its difference with other approaches. Topics include statistical analysis of input data and appropriate probability distributions, random number generation, design of simulation experiments, model accuracy and validation, variation reduction techniques, output and result sensitivity analysis. Students become familiar with modern simulation languages and learn to code simulation models using the latest simulation software. The course particularly focuses on real-life applications (eg. capacity planning and programming in manufacturing, queue management in call centers, and patient flow analysis in emergency services) that demonstrate how simulation can be used to improve processes and help decision-making in complex systems.

IE 556 Decision Support Systems                                                                           

The aim of this course is to help students develop computational and programming skills needed to build spreadsheet-based decision support tools. Topics to be covered include, pivot tables, lookup functions, VBA programming and macros, user interface development, solver and simulation model building with VBA. Students are familiarized with decision support system development for case studies in the fields of engineering, management science and finance

IE 562 Game Theory

Game theory is a mathematical tool developed to understand not only economical market participants’ interactions but also social phenomena observed as a result of these interactions. The aim of this course is to introduce the students the analytical approaches to investigate the strategic interactions of the participants.  Topics to be covered include, utility concept, games in normal form, dominance, Nash equilibrium, pure and mixed strategies, extensive form games, sequential games, games under asymmetric information. Students will be familiarized with the applications of the basic concepts in the fields of engineering, management and economics via case studies.

IE 564 Decision and Risk Analysis

The aim of this course is to teach students the analytical models used in uncertainty, conflicting goals and rational decision making at risk. Topics include modeling uncertainty, principles of rational decision making, analyzing decision problems with value trees, decision trees and impact diagrams, solving value hierarchies, decision trees and impact diagrams, defining and calculating the value of information, including risk quantity in analysis and making sensitivity analysis. The concepts and solution methods covered in the course will be presented through case studies.

IE 572 Forecasting in Management

Overview of forecasting in operations planning and control systems and demand management; data aggregation and pyramid forecasting; evaluation of the accuracy of forecasts; overview of judgmental, statistical, qualitative-based statistical, and statistical-based judgmental forecasting methods; time-series, trends, seasonality,; time-series forecasting methods (data acquisition, model identification, forecast generation costs and selecting the “best” method; effects of outliers, structural change and error evaluation choices on method selection benefits of combining alternative forecasts;; a case study in corporate demand & supply chain management.

CS 545 Deep Reinforcement Learning

This course presents introductory material for a branch of machine learning, called Reinforcement Learning, that aims to build models that learn intelligent action planning. Reinforcement learning models are at present actively used in robot control, building artificial intelligence engines of computer games, and autonomous driving. The course approaches reinforcement learning from a deep learning perspective. It also covers practical aspects of constructing deep neural network architectures that can perform reinforcement learning

CS 552 Data Science with Python

This course aims to equip graduate students with the practical Machine Learning skills to solve real-life Data Science problems using existing tools and Python programming language. The following topics will be covered during the lectures: Introduction to Python, Machine Learning with Python, Unsupervised Learning, Supervised Learning, Recommender Systems.

CS 554 Introduction to Machine Learning and Artificial Neural Networks

Course introduces and teaches standard techniques in neural networks and machine learning Topics to be covered: Concept of (machine) learning, Methods derived from mathematics (interpolation, error minimization etc.), Methods inspired from (real) neural networks of the brain (perceptron learning, Hopfield networks, feedforward networks etc.)

CS 556 Big Data Analysis

The course discusses data mining and machine learning algorithms for analyzing very large amounts of data. The emphasis is on Map Reduce as a tool for creating parallel algorithms that can process very large amounts of data. This course covers topics related to Frequent item sets and Association rules, Near Neighbor Search in High Dimensional Data, Locality Sensitive Hashing (LSH), Dimensionality reduction, Recommendation Systems, Clustering, Link Analysis, Large scale supervised machine learning, Data streams, Mining the Web for Structured Data, and Web Advertising.

CS 558 Introduction to Statistical Machine Learning               

This course covers modern statistical inference techniques from a machine learning perspective. The concept of prior knowledge is introduced and its principled incorporation into a learning model is discussed. Modern approximate inference techniques that lie at the heart of the present machine learning research are explained in detail. The course culminates with the recent developments on how these techniques can be applied to perform deep learning.

CS 566 Introduction to Deep Learning                       

This course discusses fundamental deep learning subjects such as convolutional neural networks, supervised classification, logistic regression, cross entropy. It further analyzes numerical stability of training, its importance on measuring performance. Course also applies fine tuning performance parameters and introduce various methods for fast convergence of training.