Ph.D. Candidacy Exams

Ph.D. Candidacy Examinations

All students in the Ph.D. program are required to take CSE 591 (1 credit), Research Experience in Computer Science and Engineering (course must be completed within the first two regular semesters after entering the Ph.D. program and a grade of B or higher must be achieved), and to pass a written Candidacy Examination within the first three regular semesters, which is offered at the beginning of each fall and spring semester. The examination tests the student's background preparation and problem-solving ability.

In the Candidacy Exams, students are required to attempt and pass examinations in three out of the nine subject areas listed below. Each subject area examination will have duration of two hours. By the second semester, students must have attempted at least three areas (with a demonstrated performance in all three). A student is limited to three chances to pass the candidacy, and will have to take the exam in the first semester of residence in order to avail themselves of the three-chance rule. The students will need to register with the graduate secretary for the examination and identify the three subject areas no later than two weeks before the date of the examination. Students must pass three subject area exams no later than the third regular semester after entering the Ph.D. program. Examinations taken as an M.S. or as an M.Eng. student will count towards the maximum limit of three attempts.

The level of the examination is that of undergraduate courses in the designated subject areas, although the questions are not necessarily restricted to topics covered in any specific course. The Graduate Officer of the department will assign coordinators for each examination. Special conditions set for an examination (such as open book) will be promulgated by the area coordinator four weeks prior to the examination.

Course Breadth Requirement

In addition to the candidacy exam requirement, all CSE Ph.D. graduate students will need to satisfy the following course requirements before graduation.

1. Pass a graduate breadth course from at least three of the nine subject areas.
2. Only one of these three courses can overlap with the area of the candidacy exam subject areas.
3. At least one of the two subject areas listed below must be covered either by passing the candidacy examination in that area or by completing the graduate breadth course specified.
   a. Data Structures and Algorithms
   b. Programming Languages
4. At least one of the two subject areas listed below must be covered either by passing the candidacy examination in that area or by completing the graduate breadth course specified.
   a. Operating System
   b. Computer Architecture

Note for Continuing Students admitted prior to Fall 2006

You may opt to remain in the old candidacy exam format or opt to change to test in the new format. If you opt for the new format and have passed either Exam A or Exam B previously, you will be considered to have passed two subject areas. All students opting for the new format will need to satisfy the course breadth requirements specified above.

Subject Area 1: Computational Molecular Biology:

Exam covers the following topics:

I. Molecular Biology of the Gene

•  The molecular biology of procaryotic and eucaryotic genes and genetics.
•  The biochemical basis of genetic phonomena, such as replication, repair, transcription, and translation.

II. General Biochemistry

•  Principles of the structure and function of biological molecules, including carbohydrates, lipids, membranes, proteins, and enzymes.
•  Chemistry of biological molecules; proteins and enzymes.

The following Penn State course(s) are recommended as providing a minimal background in the above topics:

BMB 400: Molecular Biology of the Gene
and BIOL 405: Molecular Evolution
or BIOL 408: Contributions of Women to the Biological Sciences Past and Present)

Suggested references:

1. Lewin. Genes V. Oxford University Press, 1994.
2. Lehninger, Nelson, and Cox. Principles of Biochemistry, 5th Edition. Worth Publishers, 2008.

BMB 400: Molecular Biology of the Gene
and BIOL 405: Molecular Evolution
or BIOL 408: Contributions of Women to the Biological Sciences Past and Present)

Subject Area 2: Computer Architecture and Organization

•  Arithmetic Logic Unit: Analysis of arithmetic operations and design of arithmetic units (adders, multipliers, dividers, etc.); floating point representations and operations.
•  CPU Design: hardware description languages; datapath design and busing structures; pipelined datapath; data and control hazards; control unit design.
•  Processor Design: instruction sets; addressing modes; performance metrics; architectural support for subroutine calls and interrupt handling.
•  Memory Systems: memory hierarchy and memory types; cache memory designs; interleaved memory; architectural support for virtual memories.
•  I/O Systems: types and characteristics of I/O devices; I/O device interfacing (I/O buses, interrupts, DMA, etc.)
•  Concurrency; Different approaches to concurrency, i.e., pipelining, parallelism, and multiprocessing, and their architectural characteristics.

The following Penn State course(s) are recommended as providing a minimal background in the above topics:

CMPEN 331: Computer Organization and Design
CMPEN 431: Introduction to Computer Architecture

Suggested references:

1. Patterson and Hennessy. Computer Organization and Design: The Hardware/Software Interface, 4th Edition. Morgan Kaufmann, 2008.
2. Hamacher et. al. Computer Organization, 5th Edition . McGraw Hill, 2002.
3. Hennessy and Patterson. Computer Architecture A Quantitative Approach, 4th Edition. Morgan Kaufmann.

CSE 530: Fundamentals of Computer Architecture

Subject Area 3: Computer Networks

Exam covers the following topics:

•  Computer Networks and the Internet: Internet, network structure, protocol stack.
•  Application Layer: HTTP, FTP, SMTP, DNS, content distribution.
•  Transport Layer: UDP, TCP, sliding window, congestion control.
•  Network Layer and Routing: Routing principles, IP, IPv6, multicast, mobile IP.
•  Link Layer and Local Area Networks: Multiple access control, LAN, Ethernet, wireless links, ATM.

The following Penn State course(s) are recommended as providing a minimal background in the above topics:

CMPEN/EE 362: Communication Networks

Suggested references:

1. Kurose, J., K. Ross.  Computer Networks: A Top-Down Approach Featuring the Internet, 5th Edition, 2010.
2. Tanenbaum, A.  Computer Networks, 4th Edition.  Prentice Hall, 2005.

Graduate Breadth Courses:

CSE 514: Computer Networks

Subject Area 4: Computer Vision and Digital Image Processing

Exam covers the following topics:

I. Digital Image Processing

•  Digital Image Fundamentals: Image sensors, sampling, quantization, pixel geometry [GW Ch.2; TV Ch.2]; Photometric concepts: color and light, illumination, reflection, sensor response [GW Ch.6; FP Ch.4&6]; Camera geometry: projection models, camera parameters, calibration [FP Ch.2; TV Ch.6].  
•  Image Enhancement: Histogram analysis and equalization. Linear operators and convolution; noise and smoothing; sharpening operators; computing image gradients; Gaussian derivatives and LoG operator; median filtering. [GW Ch.3; TV Ch.3].  
•  Frequency-Domain Filtering: Fourier transform; Discrete Fourier Transform (DFT) and FFT; filtering in frequency domain [GW Ch.4].  
•  Image Compression: Information theory and coding theorems; lossless vs. lossy compression; predictive compression methods; JPEG - still image compression; MPEG - motion video compression [GW Ch.8].  
•  Binary Morphology: Basic operations: erosion, dilation, thinning, region filling, connected components [GW Ch.9].  

II. Computer Vision  

•  Feature Extraction: Edge detection; case study: Canny algorithm [TV Ch.4]; Line/curve fitting: least squares, RANSAC and Hough transform [TV ch.5]; Corner detection; case study: Harris algorithm [TV Ch.4]; Region segmentation: split&merge; region growing [GW Ch.10].  
•  Image Mappings: Forward/inverse geometric warping and interpolation [GW Ch5.11]; Parametric transformations (translation, rotation, scale, Euclidean, similarity, affine, projective) [course notes].  
•  Stereopsis: Parallax vs. depth; correspondence problem and patch matching via SSD and NCC; epipolar geometry; Essential vs. Fundamental matrix; image rectification; stereo reconstruction [TV Ch.7; FP ch.11].  
•  Motion: Motion field vs. optical flow; aperture problem and normal flow; computing optical flow (case study: Lucas-Kanade algorithm); structure from motion; the Factorization algorithm; change detection [TV Ch.8].  
•  Tracking: Correlation-based approaches (e.g. template matching); mean-shift algorithm (e.g. CAMSHIFT) [course notes].  
•  Object Recognition: Interpretation trees; invariants; appearance-based (eigenspace) methods [TV Ch.10].  

Penn State courses CMPEN 455 and CMPEN 454 are recommended as providing a minimal background in the exam topics. It is also expected that the student has mastered background knowledge in college-level linear algebra and probability theory.

Suggested references:

1. [GW] Gonzalez and Woods.  Digital Image Processing, 3rd Edition.  Prentice Hall, 2008.
2. [TV] Trucco and Verri.  Introductory Techniques for 3D Computer Vision.  Prentice Hall, 2005.
3. [FP] Forsyth and Ponce.  Computer Vision: A Modern Approach. Prentice Hall, 2003.