EEE465 Numerical Methods for Artificial Intelligence

weekly schedule

week	part	topic
1	linear algebra	introduction; linear algebra review; block matrices and norms
2	linear algebra	linear independence and rank; subspaces and linear equations
3	least squares	least squares (LS); vector derivatives and positive semidefinite matrices
4	least squares	orthogonality and Gram-Schmidt algorithm; LS classification and cross-validation
5	singular value decomposition	motivating the singular value decomposition (SVD); matrix norms and the SVD
6	singular value decomposition	SVD geometry and principal component analysis (PCA); low-rank approximation and pseudoinverse; SVD geometry and sensitivity
7		review and preparation for the exam (midterm)
8		midterm exam
9	regularization	trade-offs and regularization; regularization examples
10	optimization for machine learning	iterative methods; proximal algorithms; convexity and support vector machine (SVM)
11	optimization for machine learning	gradient methods; stochastic gradient method; max-margin SVM and kernels
12	constrained optimization	convex optimization; bounding and duality; Karush-Kuhn-Tucker (KKT) conditions
13	machine learning	artificial neural networks (ANNs) and perceptron; convolutional neural networks (CNNs)
14	machine learning	unsupervised learning and k-means clustering; matrix problems and vectorization
15		review and preparation for the exam (final)