Computer Engineering Ph.D. Qualifying Exam Guidelines
The following guidelines are set for the Ph.D. Qualifying Examination in Computer Engineering, in addition to rules and regulations at Middle East Technical University Student Handbook.
General Information
- Ph.D. Qualifying exam consists of a written part and an oral part. The candidate is considered successful when he/she passes both parts.
- Ph.D. Qualifying exam is given twice a year each May and November.
- The candidate should get the approval of his/her advisor and petition the department at least one month before the exam. The student is required to submit the one-page abstract of her/his paper for the oral (area) exam as an attachment to her/his petition for taking the area exam.
- The candidate failing to pass the Ph.D. Qualifying exam is given a second chance in the subsequent offering of the exam. Failure in the second attempt leads to the dismissal of the student from the Ph.D. program.
Ph.D Qualifying Exam
The exam consists of two parts; Written (Core) exam and Oral (Area) exam.
Written (Core) Exam
This part of the exam covers the following 7 main topics:
- Data Structures (CENG 213),
- Algorithms (CENG 315),
- Discrete Math (CENG 223),
- Theory of Computation (CENG 280),
- Programming Languages (CENG 242),
- Operating Systems (CENG 334),
- Digital Design and Computer Architecture (CENG 232 & CENG331)
The questions from the Core part will be at the undergraduate level and will cover the content that is listed in the Core subjects table below. In the exam, there will be two questions from each topic, and the student will be asked to attempt only one question from each. Each subject is graded over 20 points and the student is considered successful if her/his total grade is 84 out of 140 (which corresponds to 60% of the maximum grade). In addition, if the student is unsuccessful in the core exam, s/he will be exempt from the subjects on which s/he scored at least 14 out of 20. In the subsequent exam, s/he will be expected to score 60% of the maximum grade from the remaining subjects only. The student reserves the right to “not be exempt” if s/he wishes. The exemption status from a course is valid only for the next exam session.
Oral (Area) Exam
The Purpose of the Exam is to evaluate the student’s ability and potential to conduct research at the doctoral level and to encourage the student to have an earlier involvement in research.
Expectations from the Student before the Exam
- To choose a topic within the student’s research field.
- To make a contribution to this topic as explained below.
- To prepare a paper in the “IEEE conference proceedings format” in at least 6 at most 8 pages.
- An Originality Report for the paper should be prepared and signed by the thesis supervisor. Similarity with the work of others shouldn’t exceed 20%. Originality Report can be prepared using Turnitin (http://www.turnitin.com) or Ithenticate (http://www.ithenticate.com) plagirism check software. It is enough to print the first page of the similarity index for the originality report.
- To submit one page abstract of this paper as an attachment to her/his petition for taking the area exam.
- To submit this paper along with the Originality Report to the examination committee (jury) at least 10 days before the first day of the oral (area) exam period.
Expectations from the Student during the Exam
- To present the student’s study ( at most 20 minutes)
- Answer questions about the study (10 minutes)
- Answer general questions in the Ph.D. field of the student (20 minutes, if the jury finds it necessary the question answer part can be extended).
Expected Contribution The student can choose to make one or more of the following types of contributions:
- Literature evaluation: A literature survey is required in every study, but those students who select this category will be expected to conduct a more detailed literature survey by identifying the advantages and disadvantages of the previous studies, compare them with each other, and provide an analysis-synthesis of the literature in the selected topic.
- Implementation: The student will implement a paper in the selected topic which should be chosen together with the student’s advisor and produce results by changing various parameter values if applicable.
- Novel approach: The student will propose a novel approach to a selected problem and implement this approach to produce results. This category includes improving an existing algorithm using a different approach.
- Comparison: The student will compare two or more algorithms that are selected with the student’s advisor, and discuss the results obtained as a result of this comparison.
- Theoretical contribution: The student will propose a novel theoretical approach such as a formula, theory, proof, etc. and show the correctness, utility, and reasoning behind this approach.
- Case study: The student will apply an existing method or process to a realistic problem and discuss the results that are obtained.
Grading The performance of the student is assessed separately by each jury member according to the table below. A grade of 60 or above is considered as a "pass" vote by that jury member. If at least three jury members vote "pass" the student is considered successful, otherwise the student fails the exam.
Ratio |
Point [0-100] |
|
%40 |
Written work |
|
%20 |
Presentation and questions related to the presentation |
|
%40 |
General questions in the selected field |
|
Weighted Total: |
100 |
General Principles about the Administration of the Exam
- For every student who will take this exam, a jury comprised of 5 people with Ph.D. degree and experts in the selected field will be formed by the qualifying exam committee. This jury will include the student’s advisor. It is crucial that the jury members read the study before the exam and prepare questions to be asked during the exam.
- Qualification Exam Committee designates one of the jury members as the chair. The chair is responsible for conducting the examination process.
- The written work prepared by the student must be original. It should not be put together by copying and pasting from the previous work.
- Jury members could check the document submitted by the student against plagiarism using http://www.ithenticate.com/ to which METU has a subscription.
- A study prepared for the master’s thesis cannot be directly used for his exam. A contribution is expected to be made during the Ph.D. studies even if the topic remains the same.
- An article (published or unpublished) written by the student as the primary author can be used for this exam. However, the same article cannot be used by more than one student, and if it was published it should not be published more than 12 months before the exam.
- In case of failure for the first time, a new topic can be chosen or the jury must indicate the expectations from the student for the second exam if the same topic will be used.
- In case the student is taking this exam for the second time from the same topic, a supporting document explaining the changes from the previous version must be provided by the student to the jury along with the latest version of the study.
Written (Core) Exam Syllabus
Course |
Topics |
Resources |
Data Structures |
Algorithm analysis for data structures |
* Mark Allen Weiss, Data Structures and Algorithm Analysis in C++ (3rd ed.), Addison Wesley, 2006 |
Lists, stacks, queues |
||
Trees |
||
Priority queues |
||
Hashing |
||
Algorithms |
Analysis of Algorithms |
* Introduction to Algorithms, T. H. Cormen, C. E. Lieserson, R. L. Rivest, C. Stein, Mc Graw-Gill |
Sorting, Searching |
||
String Processing |
||
Graph Algorithms |
||
Greedy Approach |
||
Divide and Conquer Algorithms |
||
Dynamic Programming |
||
Exhaustive Search |
||
Complexity Classes, NP-completeness |
||
Discrete Mathematics |
Propositional Logic: Logic, Equivalences |
* K.H. Rosen, Discrete Mathematics and its Applications, (Sixth Edition) McGraw-Hill, 2007. |
Predicate Logic: Predicates and Quantifiers, Nested Quantifiers, Methods of Proof |
||
Sets and Functions: Sets, Set Operations, Functions, Growth of Functions, Complexity of Algorithms |
||
Integers: Integers and Division, Integers and Algorithms |
||
Induction and Recursion: Sequences and Summations, Mathematical Induction, Recursive Definitions and Structural Induction, Recursive Algorithms |
||
Counting: Permutations and Combinations, Recurrence Relations, Solving Recurrence Relations, Generating Functions, Inclusion and Exclusion |
||
Relations: Relations and Their Properties, Representing Relations, Closure of Relations, Equivalence Relations, Partial Orderings |
||
Graphs: Int to Graphs, Graph Terminology, Representing Graphs, Connectivity, Euler and Hamiltonian Paths, Shortest Path Problem, Graph Coloring |
||
Trees: Int to Trees, Applications of Trees, Spanning Trees, Min Spanning Trees |
||
Theory of Computation |
Finite Automata and Regular Expressions: Alphabets and languages, Finite representations of languages,Deterministic finite automata, Nondeterministic finite automata, Equivalence of DFA and NFA, Finite automata versus regular languages, Pumping lemma and its applications, State minimization |
* Elements of the Theory of Computation, H.R.Lewis, C.H.Papadimitriou, (2nd ed.), Prentice-Hall, 1998. |
Push-down Automata and Context Free Grammars: Parse trees and derivations,Pushdown automata, Pushdown automata versus context-free grammars, Closure properties,Pumping theorem and its applications, Deterministic PDAs |
||
Regularity and context-freeness of languages |
||
Turing Machines and unrestricted grammars: Turing machines – definition and examples, Computing with TMs, Recursive and recursively enumerable languages, Extensions of TMs, Nondeterministic TMs, Unrestricted grammars |
||
Church-Turing thesis, universal Turing machines |
||
Halting problem |
||
Programming Languages |
Storage structures, control structures, scope and binding |
* Programming Language Concepts and Paradigms, D.A. Watt, Prentice-Hall, 1990. |
Data and procedural abstraction |
||
Type systems |
||
Lexical and syntactic description of languages |
||
Object-oriented programming languages |
||
Functional programming languages |
||
Logic programming languages |
||
Operating Systems |
Operating Systems Structures |
* Modern Operating Systems, A.S. Tanenbaum, Prentice-Hall, ISBN 0-13-595752-4, 1992. |
Processes, Threads and Their Management |
||
Process and Processor Scheduling |
||
Process Synchronization |
||
Interprocess Communication |
||
Deadlocks |
||
Memory Management |
||
Storage Management (I/O Processing, File Systems) |
||
Protection and Security |
||
Digital Design and Computer Architecture |
Combinational Circuits |
* Digital Design, M. Mano, Prentice-Hall, ISBN 0-13-212994-9, 1991. |
Combinational Circuit Minimization: Algebraic and Karnaugh-map minimization |
||
Synchronous Sequential Circuits |
||
Registers, Counters |
||
RAM, ROM, PLA, and PAL |
||
Arithmetic Logic Unit, Multiplication and Division, Floating Point operations |
||
Pipelining: Hazards, Forwarding, Branch Prediction |
||
Memory Hierarchy: Interleaving, Cache Memory, Virtual Memory |
||
I/O Systems: Buses, I/O Interfaces, Interrupts, DMA |
Oral (Area) Exam Syllabus
Course |
Topics |
Resources |
Artificial Intelligence |
Uninformed and Heuristic Search |
* Artificial Intelligence: A Modern Approach, S.Russell, P.Norvig, Prentice Hall, 1995. |
Game Playing |
||
Constraint Satisfaction and Propagation |
||
Knowledge and Reasoning |
||
Theorem Proving |
||
Planning |
||
Reasoning with Uncertainty |
||
Machine Learning: Learning from examples (supervised learning, decision trees, Regression and classification, ANN, SVM), Learning probabilistic models (Bayesian learning, Naive Bayes classifiers, EM algorithm), Reinforcement Learning (passive RL, active RL) |
||
Computer Graphics |
Rendering Pipeline: Major stages of the rendering pipeline |
* Computer Graphics: Principles and Practice, Foley, Van Dam, Feiner, Hughes, (2nd ed.), Addison Wesley, 1995. |
Geometric Transformations: Homogeneous coordinates, Vectors, points, normals, Translation, scaling, rotation, sheer transformations (2D and 3D) |
||
Raster Algorithms: Line rasterization, Triangle rasterization, Antialiasing |
||
Viewing: Parallel projections, Perspective projections, Clipping, Viewport transformation |
||
Visible Surface Detection: Back-face elimination, Z-buffer algorithm |
||
Phong Shading Model: Ambient Light, Diffuse Reflection, Specular Reflection |
||
Polygonal Surface Shading: Flat shading, Goraud shading, Phong shading |
||
Texturing: Generating of uv coordinates (for both 2D and 3D texture mapping), Mipmapping, Bilinear interpolation, Bump mapping |
||
Volume Rendering: Marching cubes algorithm, Direct volume rendering |
||
Three Dimensional Object Representations: Hermite curve, Natural cubic splines, Bezier curves and surfaces, Geometric continuities, Joining curves and surfaces |
||
Ray tracing: Parametric lines, Parametric and implicit surfaces, Ray-object intersections (triangle, sphere, plane), Basic ray tracing algorithm, Generating simple shadows with ray tracing, Accelleration structres (bounding boxes, oct-tree, kd-tree) |
||
Radiosity: Basic radiosity algorithm, Radiosity equation, Hemicube method for form factor calculations, Jacobi iteration and Gauss Seidel for solving Ax=b |
||
Natural Language Processing |
Linguistic knowledge representation and propagation |
* Speech and Language Processing, Jurafsky and Martin, Prentice-Hall, 2000. |
Computational aspects of Morphology |
||
Syntactic representation in NLP (phrase structure, dependency, unification) |
||
Parsing strategies for natural languages (bottom-up,top-down, mixed) |
||
Parsing decisions and improvements (determinism, non-determinism, charts) |
||
Grammar formalisms (dependency grammars, categorical grammars, phrase-structure grammars) and hierarchy for natural languages |
||
Handling non-local dependencies |
||
Compositional semantics: Lambda-calculus and logical form |
||
Basics of data-intensive linguistics (n-grams, language models, classifiers) |
||
Database Systems |
Physical data organization |
* Database Management Systems, Raghu Ramakrishnan, McGraw-Hill. |
Data models |
||
Relational database design theory (normalization) |
||
Relational query languages |
||
Integrity and security |
||
Transaction management |
||
Concurrency control |
||
Recovery techniques |
||
Query optimization |
||
Numerical Computation |
Numerical stability of algorithms and conditioning of problems |
* Numerical Methods, G.Dahlquist, A.Björck, Prentice-Hall. |
Linear systems: Norms, matrix norms, Gaussian elimination, forward and backward substitution, pivoting, Householder’s reflection, Given’s rotations, Gram-Schmidt method, QR, Singular Value Decomposition, Linear Least Squares problems and curve fitting, Relaxation methods (Jacobi, Gauss-Seidel) |
||
Matrix eigenvalue Problems: Power method, inverse iteration, Rayleigh Quotient, and QR iterations, Jacobi method, Arnoldi and Lanczos processes, Krylov subspace methods for solution of linear systems (GMRES, CG, BiCGStab), preconditioning |
||
Finding roots of nonlinear equations: Bisection, Secand, Newton’s methods, fixed point iteration |
||
Interpolation: Lagrange interpolation, Newton’s interpolation and divided differences, Runge’s phenomenon, Splines, Orthogonal polynomials |
||
Numerical integration: Interpolatory quadrature, Composite quadrature rules |
||
Software Engineering |
Lifecycles and process models |
* Software Engineering: a Practitioners Approach, R.S. Pressman, (4th ed.), McGraw-Hill. |
Software project management |
||
Specification and modeling techniques |
||
Traditional, object oriented and component based approaches |
||
Software metrics |
||
Software quality |
||
Testing and integration methods |
||
Maintenance |
||
Pattern Recognition and Image Analysis |
Image Transform: Discrete Fourier transform (FFT excluded), Discrete Haar Wavelet transform |
* Digital Image Processing, R. C. Gonzales and R. E. Woods, Prentice-Hall, 3rd edition, 2008. |
Image Enhancement Techniques: Point Processing (basic intensity transformations), Histogram processing, Image negation, power law, log transformations, Spatial Filtering, Convolution (smoothing, sharpening) |
||
Image Compression: Redundancy and measuring image information, Huffman coding |
||
Morphological Operations: Erosion, dilation, opening, closing |
||
Image Segmentation: Edge detection (Canny, Hough transform), Thresholding |
||
Image Representation and Description: Chain codes, Polygons, Regional descriptors |
||
Texture: Texel-based Texture Descriptions, Quantitative Texture Measures |
||
Content-based image retrieval: Image Distance Measures (Color, Texture and Shape Similarity Measures), Precision, Recall and F-score Performance Analysis |
||
Motion from 2D images: Image Subtraction |
||
Stereo Vision: Matching: Cross Correlation, Symbolic Matching, The Epipolar and The Ordering Constraints |
||
Bayesian Decision Theory: Gaussian Density Estimation, Classifier Discriminant Functions |
||
Maximum Likelihood Method: Gaussian Density Estimation |
||
Non-parametric techniques: Parzen Window, K-Nearest Neighbor |
||
Unsupervised learning: Mixture Resolving, Unsupervised Bayes Method, Maximum Likelihood Method |
||
Clustering: K-means Clustering, Hierarchical Clustering, Component Analysis |
||
Neurocomputing |
Learning and generalization |
* Neural Computing: Theory and Practice, P.D. Wasserman. |
Multilayer perceptrons and the backpropagation algorithm |
||
Hopfield model |
||
Recurrent networks |
||
Unsupervised learning and self organizing maps |
||
Adaptive resonance theory |
||
Radial basis function networks |
||
Higher order neural networks |
||
Neurodynamics |
||
Parallel Computing |
Parallelism and classification of parallel computers: Performance bottlenecks, Classification of parallel computers and applications, Programming models for parallel computers |
* Introduction to Parallel Computing, by Grama, Gupta, Kumar, and Karypis, Addison Wesley, 2003. |
Pipelining and vector processing: Instruction pipelining, superscalar execution, and instruction scheduling, Pipelining arithmetic operations, Performance analysis of pipelined operations |
||
Interconnection topologies and implementing various communication operations: Metrics for evaluating performance of interconnection networks, Point to point and collective communication operations and their implementation |
||
Task decomposition and design of parallel algorithms: Principles of parallel algorithm design, Task interaction and dependency graphs, Graph partitioning/clustering, Load balancing |
||
Analysis of parallel algorithms: Speed improvement and efficiency, Amdhal’s law, Gustafson’s law, Weak and strong scalability |
||
Parallelism in various applications (e.g. matrix problems in scientific applications, sorting and searching, etc.) |
||
Distributed Systems |
Time Synchronization |
* Distributed Systems: Principles and Paradigms, 2nd edition, A.S. Tanenbaum, M. Van Steen, Pearson Higher Education, 2007. |
Coordination |
||
Structuring Distributed Systems |
||
Process Interaction and Group Communication |
||
Distributed File Systems |
||
Concurrency Control |
||
Distributed Shared Memory |
||
Basics of Fault-Tolerance and Real-Time Systems |
||
Programming Languages and Compilers (Advanced) |
Typed lambda calculus |
* Foundations for Programming Languages, (first six chapters) J.C. Mitchell, MIT Press, 1996. |
Semantic specification of languages: Operational, denotational and axiomatic approaches |
||
Algebraic specification of data types |
||
Partial correctness proofs with before and after assertions |
||
Lexical and syntactic analysis of languages |
||
Syntax-directed translation, attribute grammars |
||
Abstract machines, intermediate languages |
||
Code generation |
||
Networked Systems |
The principles and techniques employed in computer and wireless networks; the seven-layer protocol suite known as ISO model |
* Computer Networking: A top down approach, 6th Ed., J.F. Kurose, K.W. Ross, Addison-Wesley, 2012. |
Data link layer issues (medium access control, reliable data transfer) |
||
Network layer issues (packet- versus circuit-switching, routing algorithms, IP, QoS) |
||
Transport layer issues (error control, flow control, congestion control, end-to-end argument, TCP, UDP) |
||
Network programming (socket interface) |
||
Performance evaluation of computer networks |
||
Security of computer networks (confidentiality, integrity, and authentication) |
||
Wireless networks (Cellular networks, mobility management, WLAN) |
||
Bioinformatics |
Sequence analysis, next generation sequencing: Genome annotation, Computational evolutionary biology, Comparative genomics, Genetics of disease, Analysis of mutations in cancer |
* Understanding Bioinformatics, M. Zvelebil and J.O. Baum, Garland Science, 2008. |
Gene and protein expression, gene regulation |
||
Structural bioinformatics: Protein folding problem, prediction of secondary/tertiary structure, Structural alignment, Multiple structural alignment, Protein docking |
||
Functional classification of proteins, human genome annotation |
||
Statistical modeling of biological data |
||
Biological Text Mining |
||
Bioimage Informatics: High-throughput image analysis |
||
Biological networks and computational systems biology |