Veri Bilimi ve Yapay Zeka Enstitüsü

Director: Şefik Şuayb Arslan

Vice Directors: Hüseyin Oktay Altun

Associate Professors: Ercan Atam, Şener Özönder

Assistant Professors: Hüseyin Oktay Altun, Mustafa Taha Koçyiğit, Şaziye Betül Özateş

Instructors:

Assistants: Ahmet Bilal Arıkan, Emre Fişne, Nesibe Şebnem Paluluoğlu, Bahri Atakan Yıldız, Atakan Zeybek, Zeynep Ellialtı, Şeyma Nefise Satıcı, Havva Eda Körpe

*Part-time

The Institute for Data Science and Artificial Intelligence is committed to being a world-class research and education center at Boğaziçi University on data science and artificial intelligence fields. The institute will promote a collaborative environment that facilitates the convergence of academic and industry partners to address a diverse range of complex challenges spanning multiple domains, including engineering, social and life sciences, business, and medicine.

The Institute offers interdisciplinary MSc and PhD programs that provide students with a rigorous academic foundation and advanced research capabilities in the fields of data science and artificial intelligence. The Institute's programs provide a comprehensive education that combines theoretical and practical coursework and research projects to equip graduate students with the skills and knowledge required to propose innovative techniques and to apply the latest technology in data science and artificial intelligence to solve real-world problems accurately. These programs are designed to train researchers/academics who possess the expertise and knowledge to make significant contributions to the digital transformation through cutting-edge research.

MASTER OF SCIENCE PROGRAM IN DATA SCIENCE AND ARTIFICIAL INTELLIGENCE

The Master of Science Program in Data Science and Artificial Intelligence is designed to prepare students for a professional career in artificial intelligence and/or data science in applied settings, as well as providing a solid basis for continued advanced research. Introductory level computer programming, probability, statistics, and mathematics skills are required for a student to start the program. It provides basic knowledge about the field through the mandatory courses: Data Science, Machine Learning, and Statistical Inference. While the program offers a thorough grounding in various aspects of data science and artificial intelligence, it also offers scientific excellence in highly interdisciplinary fields that aim to solve important societal problems.

The MS Program in Data Science and Artificial Intelligence is an interdisciplinary program. Therefore, people from diverse backgrounds like engineering, law, medicine, life, and social sciences are all candidates for the program. The program includes courses from different research fields of data science and artificial intelligence which can address needs of a wide spectrum of students.

This is a two-year program with the first year devoted to completing all the course requirements, and the second year devoted to preparation of a thesis based on authentic research. For students lacking the required scientific background, an additional one-year remedial program is offered.

Table 1: The Master of Science Curriculum in Data Science and Artificial Intelligence

First Semester		Credit	ECTS
DSAI 510	Data Science	4	8
DSAI 512	Machine Learning	4	8
DSAI 514	Statistical Inference	4	8
— —	Complementary Elective	3	6
Subtotal:		15	30

Second Semester		Credit	ECTS
DSAI —	Area Elective*	3	7
DSAI —	Area Elective*	3	7
— —	Complementary Elective	3	6
DSAI 579	Graduate Seminar	0	2
DSAI 599	Guided Research	0	8
Subtotal:		9	30

		Credit	ECTS
DSAI 690	Master's Thesis	0	60
Program total:		24	120

*All area elective courses must be chosen from the courses listed in Table 4.

Table 2: Remedial Courses

First Semester		Credit	ECTS
DSAI 301	Introduction to Programming with Python	4	8
DSAI 303	Probability and Statistics for Data Science and AI	4	8
Subtotal:		8	16

Second Semester		Credit	ECTS
DSAI 302	Python for Data Science and AI	4	8
DSAI 304	Mathematics for Data Science and AI	4	8
Subtotal:		8	16

Candidates with gaps in their background may be required to complete a remedial program before they start the program. Remedial courses are listed in Table 2, nevertheless; each candidate is responsible for a specific subset of courses based on his/her scientific background. The remedial program must be completed within one or two semesters depending on the admission requirements.

Data Science and Artificial Intelligence M.S. Program includes three compulsory courses: DSAI510 Data Science, DSAI512 Machine Learning, and DSAI514 Statistical Inference. Each student is also required to take at least two area elective courses, two complementary elective courses, a graduate seminar course, and a guided research course.

Each student is required to choose a specific area associated with his/her studies. The program offers two fundamental specialization areas: Data Science and Artificial Intelligence. The “Area Elective Courses” of two specialization areas of the program are provided in Table 4. These courses must be approved by the student’s academic or thesis advisor. The seminar course is designed to expand the research perspective of the prospective students. The guided research course is given by the thesis advisors to specify student’s research direction and thesis proposal.

Individuals registered for the program are required to choose their thesis advisor and research field until the end of the first semester. They must prepare a thesis proposal and must submit the thesis title to the Institute until the end of the second semester.

DOCTOR OF PHILOSOPHY PROGRAM IN DATA SCIENCE AND ARTIFICIAL INTELLIGENCE

Doctor of Philosophy program in Data Science and Artificial Intelligence provides a regular PhD program for candidates who have an master’s degree in a program related to data science or artificial intelligence and an integrated PhD program for candidates who have only a bachelor’s degree or a master’s degree from other programs.

The students in the integrated PhD program are required to accumulate a minimum of 42 credits by completing at least 14 graduate courses three of which must be the compulsory courses of the MS and the regular PhD program in Data Science and Artificial Intelligence. They may also be asked, based on their background, to complete the remedial course work prior to starting the integrated PhD program. The remedial courses are given in Table 2. The remedial program must be completed within one or two semesters depending on the admission requirements.

The PhD program is given in Table 3. In the program, individuals are required to determine a specific curriculum with their advisors to guide them for their research interests. Each student is required to take a graduate seminar course in which all research fields of the program are introduced, a guided research course, at least four Area Elective Courses selected from Table 4, and at least three complementary elective courses which may be chosen from graduate courses offered by the Institute or the graduate level courses of other programs based on the student’s research field. The guided research course is given by the thesis advisors to specify student’s research direction and thesis proposal prior to the qualifying exam.

Individuals must prepare a thesis proposal while they are taking the guided research course. The PhD thesis is required to be completed in the legal term for every individual registered in the program following the approval of the thesis proposal.

Table 3: The Doctor of Philosophy Curriculum in Data Science and Artificial Intelligence

First Semester		Credit	ECTS
DSAI —	Area Elective*	3	8
DSAI —	Area Elective*	3	8
— —	Complementary Elective	3	7
— —	Complementary Elective	3	7
Subtotal:		12	30

Second Semester		Credit	ECTS
DSAI —	Area Elective*	3	7
DSAI —	Area Elective*	3	7
— —	Complementary Elective	3	6
DSAI 700	Graduate Seminar	0	2
DSAI 699	Guided Research	0	8
Subtotal:		9	30

		Credit	ECTS
—	Qualifying Exam	0	30
—	Thesis Proposal Defense	0	30
DSAI 790	PhD Thesis	0	120
Program total:		21	240

*All area elective courses must be chosen from the courses listed in Table 4.

Table 4: DSAI All Area Elective Courses of the MS and PhD Programs in Data Science and Artificial Intelligence

DSAI 511	Algorithms
DSAI 520	Big Data Systems
DSAI 521	Data Visualization for Data Scientists
DSAI 522	Business Intelligence & Analytics
DSAI 523	Cloud Computing & Distributed Systems
DSAI 524	Software Design for Data Science
DSAI 525	Time Series & Forecasting with ML
DSAI 526	Web Mining
DSAI 530	Foundations of Computational Social Science
DSAI 531	Social Media Analytics
DSAI 532	Digital Humanities
DSAI 533	Human-Centered Systems
DSAI 540	Theory of Computational Intelligence
DSAI 541	Deep Learning
DSAI 542	Reinforcement Learning
DSAI 543	Image Processing with ML
DSAI 544	Computer Vision with ML
DSAI 545	Natural Language Processing
DSAI 546	Heuristic Optimization
DSAI 549	Ethics, Policy, Governance & Regulation in AI
DSAI 550	Introduction to Cognitive Science
DSAI 551	Data-Driven Modelling & Control
DSAI 591	Directed Studies I
DSAI 592	Directed Studies II
DSAI 641	Advanced Machine Learning
DSAI 642	Advanced Reinforcement Learning
DSAI 643	Meta-Learning
DSAI 644	Graph Neural Networks
DSAI 645	Optimization for AI
DSAI 651	Dynamic System Modelling
DSAI 652	Autonomous Vehicles
DSAI 691	Directed Studies I
DSAI 692	Directed Studies II

COURSE DESCRIPTIONS: DSAI 301 Introduction to Programming with Python (4+1+0) 4 ECTS 8

(Python ile Programlamaya Giriş)

Introduction to programming. Jupyter notebook. Basic data types. Statements, blocks. Selection structures, repetition structures, recursion. Functions. Strings. Input/Output and files. Python data types.

Prerequisite: -

DSAI 302 Python for Data Science and Artificial Intelligence (4+1+0) 4 ECTS 8

(Veri Bilimi ve Yapay Zeka için Python)

Advanced data types. Iterators and generators. Arrays with Numpy. Data frame structures, input-output, and data analysis with Pandas. Data visualization with Matplotlib. Scientific computing, optimization, and machine learning with SciPy and TensorFlow.

Prerequisite: -

DSAI 303 Probability and Statistics for Data Science and Artificial Intelligence (4+1+0) 4 ECTS 8

(Veri Bilimi ve Yapay Zeka için Olasılık ve İstatistik)

Basic probability models. Discrete and continuous random variables. Data analysis using statistical tools. Random sampling. Regression and estimation. Statistical methods for data science and machine learning. Random processes.

Prerequisite: -

DSAI 304 Mathematics for Data Science and Artificial Intelligence (4+1+0) 4 ECTS 8

(Veri Bilimi ve Yapay Zeka için Matematik)

Functions of multi-variables and gradient. Fundamentals of linear algebra relevant for data science and artificial intelligence. Basics of differential equations. Basic optimization for data science and artificial intelligence. Essential information theory concepts and results relevant for data science and artificial intelligence.

Prerequisite: -

DSAI 510 Data Science (4+1+0) 4 ECTS 8

(Veri Bilimi)

Data types: structured, unstructured, streaming. Data formats. Computer programming for data science. Exploratory data analysis, data cleaning, imputing and transforming data, data visualization, feature extraction and dimensionality reduction, knowledge discovery. Applications of supervised, unsupervised and semi-supervised machine learning for data science. Model evaluation, performance metrics. Data scrapping, text mining, data ethics and data bias.

Prerequisite: -

DSAI 511 Algorithms (3+1+0) 3 ECTS 8

(Algoritmalar)

Basic data structures. Basic algorithms: searching and sorting. Performance analysis of algorithms. Algorithm design techniques: divide and conquer, greedy, dynamic programming, branch and bound. Heuristic algorithms and their analysis. Randomized algorithms.

Prerequisite: -

DSAI 512 Machine Learning (4+1+0) 4 ECTS 8

(Makine Öğrenmesi)

Introduction to machine learning. Application areas of machine learning. Machine learning paradigms: supervised, unsupervised, semi-supervised and reinforcement learning. Evaluation of machine learning models. Classification and clustering. Dimensionality reduction. Regression models. Artificial neural networks. Support vector machines. Kernel methods. Decision trees. A brief overview of some advanced machine learning techniques.

Prerequisite: -

DSAI 514 Statistical Inference (4+1+0) 4 ECTS 8

(İstatistiksel Çıkarsama)

Randomization and sampling, probability models, data visualization, simple and multiple regression, bootstrapping, confidence intervals, hypothesis testing, inference for regression, correlation, and causation.

Prerequisite: -

DSAI 520 Big Data Systems (3+1+0) 3 ECTS 8

(Büyük Veri Sistemleri)

Introduction to Unix and shell scripting. Traditional and modern big data systems. Distributed file systems and data storage, streaming data handling, query languages (SQL and NoSQL), distributed data processing, big data extraction and integration, system optimization. Hadoop ecosystem. Big data applications (spatial-temporal, multimedia, health, social media, and scientific data): analysis, visualization, machine learning workflow. Basics of cloud computing.

Prerequisite: -

DSAI 521 Data Visualization for Data Scientists (3+1+0) 3 ECTS 8

(Veri Bilimciler için Veri Görselleştirme)

Visualization of different data types (plots, charts, maps, trees). Static and interactive data publishing. Commercial data analysis and presentation tools. Open-source tools. Designing and developing dashboards. Publishing data on the Internet. Storytelling with data.

Prerequisite: -

DSAI 522 Business Intelligence and Analytics (3+1+0) 3 ECTS 8

(İş Zekası ve Analitiği)

Overview of data storing, dashboarding and data warehousing. Data handling and querying to discover patterns. Creating custom data visualizations. Practical tools for analyzing data. Techniques to identify errors and trends in the data. Development of prediction models. Data-informed decision making. Data interpretation and storytelling.

Prerequisite: -

DSAI 523 Cloud Computing and Distributed Systems (3+1+0) 3 ECTS 8

(Bulut Bilişim ve Dağıtık Sistemler)

Introduction to distributed systems. Distributed programming models: inter-process communication, remote invocation, web services. Scalable system design with cloud computing. Virtualization. Data replication and availability. Data lakes, data warehouses and OLAP. Cloud programming and software environments. Cloud services. Data pipelines and machine learning on the cloud.

Prerequisite: -

DSAI 524 Software Design for Data Science (3+2+0) 3 ECTS 8

(Veri Bilimi için Yazılım Tasarımı)

Software engineering principles. Software life cycle: requirements elicitation, design, development, testing, deployment, maintenance. DevOps methodology and tools: agile development, component design, integration/continuous deployment, version control, build automation. MLOps fundamentals: data drift, reproducibility, containerization.

Prerequisite: -

DSAI 525 Time Series and Forecasting with Machine Learning (3+1+0) 3 ECTS 8

(Makine Öğrenmesi ile Zaman Serileri ve Öngörü)

Time series data formats and databases, cleaning timestamped data. Exploratory data analysis. Statistical analysis of time series data and outlier detection. Visualization. Autoregressive models. Seasonality in data. Autocorrelation function, smoothing, Kalman filtering. Machine learning for time series data. Deep learning with LSTM and RNN. Time series as a regression problem. Forecasting with models, anomaly detection. Reinforcement learning with time series data. Streaming data on the cloud. Time series applications.

Prerequisite: -

DSAI 526 Web Mining (3+1+0) 3 ECTS 8

(Web Madenciliği)

Web crawlers and Web scrapping. Web content mining: text mining, social media mining, sentiment analysis and opinion mining. Web usage/traffic mining and metrics. Feature selection and machine learning in text mining.

Prerequisite: -

DSAI 530 Foundations of Computational Social Science (3+1+0) 3 ECTS 8

(Hesaplamalı Sosyal Bilimin Temelleri)

Introduction to computational social science and social research. Algorithms and society. Experimental designs, randomized control trials. Text as data: introduction to natural language processing (NLP). NLP methods. From data to conclusions: validity and generalizability. Networks, using networks for social science, behavior in social networks. Social dynamics: feedback in social environments.

Prerequisite: -

DSAI 531 Social Media Analytics (3+1+0) 3 ECTS 8

(Sosyal Medya Analitiği)

Visualizing and modeling patterns in social media data. Social media: text, image, video. Monitoring customer engagement in social media, sentiment analysis and topic modeling. Social media networks. Types of networks and their representation. Network and subnetwork formation. Power in networks: centrality, hierarchy, and balance. Social network analysis and metrics.

Prerequisite: -

DSAI 532 Digital Humanities (3+1+0) 3 ECTS 8

(Dijital Beşerî Bilimler)

Text as data. Digitization, image analysis and OCR. Corpus construction: data conversion, indexing, and tagging. Dictionary methods. Text mining. Visualization. Methods for natural language processing. Modes of content analysis. Frequency analysis. Sentiment analysis. Topic modeling. Word embeddings. Causal inference and prediction. Concept mapping. Archival data.

Prerequisite: -

DSAI 533: Human-Centered Systems (3+1+0) 3 ECTS 8

(İnsan Merkezli Sistemler)

History of general interfaces and computing interfaces. Properties of complex systems. Cognitive and physical characteristics of human, vision, memory, and motor/sensory systems. Visual communication principles. Conceptual design, affinity diagramming, contextual inquiry. Interface design and prototyping. Interface evaluation. Experimental design. Universal, multilingual, and multicultural design principles. Human computer interaction standards and regulations.

Prerequisite: -

DSAI 540 Theory of Computational Intelligence (3+1+0) 3 ECTS 8

(Hesaplamalı Zeka Teorisi)

Introduction to computational intelligence: computational intelligence paradigms and history of computational intelligence. Foundations of neural computation: single and multi-layer neural networks, backpropagation, radial-basis function neural networks, recurrent neural networks. Fuzzy systems: introduction to fuzzy computation, fuzzy set theory, fuzzy logic and reasoning, fuzzy clustering and classification. Evolutionary computation: introduction to evolutionary computation, evolutionary optimization, evolutionary learning, collective intelligence and other extensions of evolutionary computation. Hybrid intelligent methods: examples of hybrid intelligence systems combining various categories of computational intelligence.

Prerequisite: -

DSAI 541 Deep Learning (3+1+0) 3 ECTS 8

(Derin Öğrenme)

Artificial neural networks, stochastic gradient descent and backpropagation, optimizers, autoencoders, generative adversarial networks, convolutional neural networks, recurrent neural networks, long short-term memory (LSTM), transfer learning. Applications in tabular data, natural language processing, computer vision, audio analysis, and reinforcement learning.

Prerequisite: -

DSAI 542 Reinforcement Learning (3+1+0) 3 ECTS 8

(Pekiştirmeli Öğrenme)

Introduction to reinforcement learning. Multi-arm bandits. Markov decision processes. Tabular-based solutions: dynamic programming, Monte Carlo, temporal-difference. Eligibility traces. Planning and learning with tabular methods. Function approximation solutions (deep Q-networks). Policy approximation. Multi-agent learning. Hierarchical reinforcement learning. Partial observable environments.

Prerequisite: -

DSAI 543 Image Processing with Machine Learning (3+1+0) 3 ECTS 8

(Makine Öğrenmesiyle Görüntü İşleme)

Basics of image analysis: filtering, denoising, edge detection, image segmentation. Convolutional neural networks. Generative adversarial network (GAN). Statistical image analysis. Image features. Data augmentation. Medical image analysis. Classification. Video processing and prediction. Image processing libraries.

Prerequisite: -

DSAI 544 Computer Vision with Machine Learning (4+1+0) 4 ECTS 8

(Makine Öğrenmesiyle Bilgisayarla Görme)

Basics of digital image processing and traditional computer vision methods. Convolutional neural network architectures. Object detection and tracking. Stereo vision. Semantic segmentation. 3D reconstruction. Reinforcement learning with visual input. Embedded hardware systems for vision. Autonomous driving. Recent trends.

Prerequisite: -

DSAI 545 Natural Language Processing (3+1+0) 3 ECTS 8

(Doğal Dil İşleme)

Overview of classical natural language processing (NLP) techniques. Language models. Morphology and challenges in Turkish texts. Word representations and contextual representations. Parsing and dependency parsing. Neural machine translation. Machine comprehension. Question answering. Applications of NLP such as text classification, sentiment analysis.

Prerequisite: -

DSAI 546 Heuristic Optimization (3+1+0) 3 ECTS 8

(Sezgisel Eniyileme)

Introduction to optimization. Problem definition: problem types, static and dynamic problems. Heuristic techniques: constructive, iterative improvement, local search, and branch-and-bound. Metaheuristics techniques: trajectory-based methods such as simulated annealing and tabu search, population-based methods such as genetic algorithms and ant colony optimization. Hyper-heuristics. Performance evaluation. Machine learning for optimization.

Prerequisite: -

DSAI 549 Ethics, Policies, Governance and Regulations in Artificial Intelligence (3+2+0) 3 ECTS 8

(Yapay Zekada Etik, Politikalar, Yönetişim ve Düzenlemeler)

Definition of artificial intelligence. Artificial intelligence and ethics. Artificial intelligence principles. Trustworthy and responsible artificial intelligence. Council of Europe and CAI, EU Artificial Intelligence Act. Risk-based approach. Artificial intelligence and privacy. Artificial intelligence and responsibility. Use of artificial intelligence in the public sector, standards, and certifications.

Prerequisite: -

DSAI 550 Introduction to Cognitive Science (3+1+0) 3 ECTS 8

(Bilişsel Bilime Giriş)

Introduction to cognitive science. Physiology of the brain. Philosophy of the mind. Fundamental approaches in cognitive science: symbolic, modular, and connectionism. Development and learning. Perception and action. Memory. Language: philosophy and neurophysiology of language, language acquisition. Cognitive neuroscience and evolution of cognition. Audition and speech. Reasoning, judgment, and decision-making. Emotion. Consciousness. Artificial intelligence approaches: the computer as a cognitive entity and embedded intelligence. Human-computer interaction and robotics. Applications of cognitive science.

Prerequisite: -

DSAI 551 Data-Driven Modelling and Control (3+1+0) 3 ECTS 8

(Veriye Dayalı Modelleme ve Kontrol)

Basic system modeling and system control concepts. Overview of linear systems and linear control. Overview of nonlinear systems. Data-driven modelling: system identification and machine learning approaches for dynamical systems modelling. Modelling performance evaluation. Control-oriented modeling and model-order reduction. Overview of dynamic programming and optimal control. Predictive control. Robust stability and performance. Data-driven fault detection. Data-driven monitoring and safety control. Data-driven scheduling and planning. Physics-informed machine learning.

Prerequisite: -

DSAI 579 Graduate Seminar (0+1+0) 0 ECTS 2 P/F

(Lisansüstü Seminer)

Seminars offered by faculty members, guest speakers, and/or graduate students designed to widen the perspectives of the students on specific topics of interest and to expand their range of scientific research techniques and publication ethics.

Prerequisite: -

DSAI 581-589 Special Topics (3+1+0) 3 ECTS 8

DSAI 58A-58Z Special Topics (3+1+0) 3 ECTS 8

DSAI 59B-59Z Special Topics (3+1+0) 3 ECTS 8

(Özel Konular)

Any course offered for credit related to the data science or artificial intelligence area. The subject matter or content of the course may vary.

Prerequisite: -

DSAI 591 Directed Studies I (3+2+0) 3 ECTS 8

(Yönlendirilmiş Çalışmalar I)

DSAI 592 Directed Studies II (3+2+0) 3 ECTS 8

(Yönlendirilmiş Çalışmalar II)

Study of selected advanced topics under the supervision of one or more faculty members.

Prerequisite: -

DSAI 599 Guided Research (0+4+0) 0 ECTS 8 P/F

(Yönlendirilmiş Araştırmalar)

Research in the field of data science and artificial intelligence, by arrangement with members of the faculty; guidance of graduate students towards the preparation and presentation of a research proposal.

Prerequisite: -

DSAI 59A Guided Research II (0+4+0) 0 ECTS 8 P/F

(Yönlendirilmiş Araştırmalar II)

Continued research in the field of data science and artificial intelligence supervised by members of the faculty; guidance of graduate students towards the preparation and presentation of a research proposal.

DSAI 690 MS Thesis (0+0+0) 0 ECTS 60 P/F

(Yüksek Lisans Tezi)

An in-depth investigation of a special topic related to the research area of the student in the data science or artificial intelligence with the goal of an original contribution. Preparation and defense of an MS thesis.

Prerequisite: -

DSAI 641 Advanced Machine Learning (3+1+0) 3 ECTS 8

(İleri Makine Öğrenmesi)

Feature selection and generation. Data transformation/augmentation techniques and adjustment/calibration of model parameters. Dimensionality reduction techniques. Spectral and kernel methods. Bayesian machine learning. Decision trees and random forests. Mixture models. Ensemble learning. Online machine learning. Current challenges in machine learning.

Prerequisite: -

DSAI 642 Advanced Reinforcement Learning (3+1+0) 3 ECTS 8

(İleri Pekiştirmeli Öğrenme)

Topology of metric spaces, contractive dynamic programming, fixed point theorems and convergence, value-based reinforcement learning, developing iterative reinforcement learning algorithms, deep reinforcement learning, advantage actor critique, trust region methods, proximal policy optimization, generalized advantage estimation, trust region policy optimization.

Prerequisite: -

DSAI 643 Meta-Learning (3+1+0) 3 ECTS 8

(Meta-Öğrenme)

Learning to learn. Transfer learning. Model selection and tuning as meta-learning. Learning from almost no data: meta-interpretive learning principles. Multi-task learning. Automated machine learning. Online and continual learning. Domain adaptation and generalization. Selected applications of meta-learning. Recent advancement of meta-learning and next steps.

Prerequisite: -

DSAI 644 Graph Neural Networks (3+1+0) 3 ECTS 8

(Çizge Sinir Ağları)

Graphs and graph structured data, graph embeddings, supervised and unsupervised graph learning, graph neural networks, social networks, applications of graph neural networks in natural language processing, computer vision and basic sciences.

Prerequisite: -

DSAI 645 Optimization for Artificial Intelligence (3+1+0) 3 ECTS 8

(Yapay Zeka için Eniyileme)

Convex optimization and gradient descent. Sub gradients and stochastic gradient descent. Online machine learning. Accelerated methods. Non-convex optimization and avoiding saddle points. Min-max optimization and generative adversarial network (GAN). Dynamic programming. Reinforcement learning, and large-scale optimization.

Prerequisite: -

DSAI 651 Dynamic System Modelling (3+1+0) 3 ECTS 8

(Dinamik Sistemlerin Modellenmesi)

Introduction to dynamic models. Types of dynamic models: white-box, gray-box, and black-box models. Continuous, discrete and hybrid models. Linear, nonlinear, and chaotic models. Model linearization. Deterministic models. Stochastic models. Modeling in time and frequency domains. Markov decision processes. Monte Carlo models and simulations. Control-oriented modeling and model-order reduction. Stability analysis. Multi-model-based dynamic system modeling.

Prerequisite: -

DSAI 652 Autonomous Vehicles (3+2+0) 3 ECTS 8

(Otonom Araçlar)

Functional architecture, main subsystems. Sensors, models, and representations. Computer vision: filters, feature extraction, line detection, projections. Motion modelling: coordinate frames and transforms, point mass model. Navigation and planning, mission planning, motion planning. Nonlinear filtering and state estimation, Bayes filter, Kalman filter, particle filter, simultaneous localization, and mapping (SLAM). Object tracking. Behavioral planning. Reinforcement learning. Vehicle control.

Prerequisite: -

DSAI 681-689 Special Topics (3+1+0) 3 ECTS 8

DSAI 68A-68Z Special Topics (3+1+0) 3 ECTS 8

DSAI 69E-69Z Special Topics (3+1+0) 3 ECTS 8

(Özel Konular)

Any course offered for credit related to the data science or artificial intelligence area. The subject matter or content of the course may vary.

Prerequisite: -

DSAI 691 Directed Studies I (3+2+0) 3 ECTS 8

(Yönlendirilmiş Çalışmalar I)

DSAI 692 Directed Studies II (3+2+0) 3 ECTS 8

(Yönlendirilmiş Çalışmalar II)

Study of selected advanced topics under the supervision of one or more faculty members.

Prerequisite: -

DSAI 699 Guided Research (0+4+0) 0 ECTS 8 P/F

(Yönlendirilmiş Araştırmalar)

Research in the field of data science and artificial intelligence, by arrangement with members of the faculty; guidance of doctoral students towards the preparation and presentation of a research proposal.

Prerequisite: -

DSAI 69A Guided Research II (0+4+0) 0 ECTS 8 P/F

(Yönlendirilmiş Araştırmalar II)

Continued research in the field of data science and artificial intelligence supervised by members of the faculty; guidance of doctoral students towards the preparation and presentation of a research proposal.

DSAI 69B Guided Research III (0+4+0) 0 ECTS 8 P/F

(Yönlendirilmiş Araştırmalar III)

Continued research in the field of data science and artificial intelligence supervised by members of the faculty; guidance of doctoral students towards the preparation and presentation of a research proposal.

DSAI 69C Guided Research IV (0+4+0) 0 ECTS 8 P/F

(Yönlendirilmiş Araştırmalar IV)

Continued research in the field of data science and artificial intelligence supervised by members of the faculty; guidance of doctoral students towards the preparation and presentation of a research proposal.

DSAI 69D Guided Research V (0+4+0) 0 ECTS 8 P/F

(Yönlendirilmiş Araştırmalar V)

Continued research in the field of data science and artificial intelligence supervised by members of the faculty; guidance of doctoral students towards the preparation and presentation of a research proposal.

DSAI 700 Graduate Seminar (0+1+0) 0 ECTS 2 P/F

(Lisansüstü Seminer)

Seminars offered by faculty members, guest speakers, and/or graduate students designed to widen the perspectives of the students on specific topics of interest and to expand their range of scientific research techniques and publication ethics.

Prerequisite: -

DSAI 790 PhD Thesis (0+0+0) 0 ECTS 120 P/F

(Doktora Tezi)

An in-depth investigation of a topic related to the research area of the student in the data science and artificial intelligence field with the goal of an original contribution. Preparation and defense of a PhD thesis.

Prerequisite: -