### GRADUATE COURSES (from Sabanci University)

**CS512:Machine Learning:-**Concept learning with version spaces, learning decision trees, statistical learning methods, Bayesian learning methods, explanation-based learning, and reinforcement learning. Theoretical aspects such as inductive bias, the probably approximately correct learning, and minimum description length principle.**EE550:Random Processes:-**Random processes and sequences, stationarity and ergodicity properties of auto- and cross-correlation functions, white noise, power spectral density and spectral estimation simulation of random processes, whitening, linear and non-linear estimation, and Wiener filtering.**EE555:Wireless and Mobile Networks**:- Principles of air interface design, characteristics of the wireless medium, wireless medium access alternatives, wireless network planning and cellular design, mobility management, and applications in wireless wide area networks, including first and second- generation mobile systems and associated networks (GSM, IS-54,IS-95), third generation wireless network (W-CDMA), wireless local area networks (IEEE 802.11 HIPERLAN),wireless ad hoc networks.**EE560:Advanced Digital Signal Processing**:- Time-frequency representations, filter banks/wavelets, multirate and polyphase filters, linear prediction, inverse problems including least squares, LMS, SVD and reconstruction from projections, adaptive filtering, non-recursive and iterative search techniques with examples on optimal quantization, Lloyd-Max quantizers and vector quantization, multi-dimensional signal processing.**EE562:Digital Speech Processing**:- Segmental descriptions of speech, the vocal mechanism, digital models for speech production, digital waveform coding, time-domain analysis methods, differential, predictive, and adaptive quantization, short-time spectrum analysis, linear prediction analysis (LPC) methods, pitch detection and vocoders, analysis-by-synthesis systems modern coding techniques and standards. Fundamentals vocoders, analysis-by-synthesis systems, modern coding techniques and standards. Fundamentals of speech recognition, dynamic time warping, and Hidden Markov Models (HMM).**EE563:Digital Image Processing**:- Digital Image Processing Imaging modalities and application areas, the electromagnetic spectrum. Two-dimensional sampling, aliasing, and quantization. Image representation, unitary transforms. Image enhancement, point operations, histogram processing, filtering. Image restoration and reconstruction, image deblurring, inverse problems, computed tomography. Image segmentation, pixel-based, edge-based, and region-based techniques, active contours. Image compression. Pattern recognation and scene interpretation.**EE566:Pattern Recognition**:- Statistical Pattern Recognition: Parameter Estimation and Supervised Learning, Bayesian Decision Theory, nonparametric approaches (Parzen windows, Nearest Neighbor), Linear Discriminant Functions, Feature extraction/selection; Pattern Recognition via Neural Networks; Syntactic Pattern Recognition; Nonmetric Methods, Unsupervised Learning and Clustering, Hidden Markov Models, Classifier Combination.**EE568:Detection and Estimation Theory**:- Principle of estimation, detection and time series analysis. Estimation: Linear and nonlinear minimum mean squared error ,estimation and other strategies. Detection: simple, composite, binary and multiple hypotheses, Neyman-Pearson and Bayesian approaches. Time series analysis: Wiener, Kalman filtering , prediction and modal Analysis.**EE571:Linear Systems**:- Gives the fundamental theory of linear dynamical systems in both continuous and discrete time, state- space representations, vector spaces, linear operators, eigenvalues and eigenvectors, functions of vectors and matrices, solutions to state equations, stability, controllability, observability, realization theory, feedback and observers.**ENS525:Mathematical Methods for Scientists and Engineers I**:- Linear vector spaces: Inner products, linear operators, eigenvalue problems, functions of operators and matrices, Fourier transforms, Hilbert spaces, classical orthogonal polynomials, Fourier series, Bessel functions, and partial differential equations.**IE509:Nonlinear Programming**:- Review on linear algebra and analysis, convex sets and functions, quadratic programming, descent algorithm, line search, conjugate directions, Newton’s method, optimization of nondifferentiable functions, necessary and sufficient conditions for constrained optimization problems, duality theory, penalty and barrier methods, Kuhn-Tucker methods, introduction to semi-infinite and semidefinite optimization, applications.

### ONLINE COURSES

**Probabilistic Graphical Models**:- (Stanford University via coursera)

Topics covered include:

- The Bayesian network and Markov network representation, including extensions for reasoning over domains that change over time and over domains with a variable number of entities
- Reasoning and inference methods, including exact inference (variable elimination, clique trees) and approximate inference (belief propagation message passing, Markov chain Monte Carlo methods)
- Learning parameters and structure in PGMs
- Using a PGM for decision making under uncertainty.