Planned Topics

• Introduction and prerequisites refresher: Course goals, linear algebra and probability

• Representation and interpretation of signals: Human perception of signals, sampling, quantization, the frequency domain, image and sound representations

• Feature analysis and discovery: Useful fixed transforms (DCT, etc), adaptive transforms (KLT/PCA/EM-PCA/online-PCA), feature extraction from familiar signals (audio, video), eigenfaces

• More feature analysis and dimensionality reduction: Independent Component Analysis (ICA), Non-Negative Matrix Factorization (NMF), Kernel PCA, Manifold embedding methods, random projections

• Detection and classification: Matched filters, template matching, object detection, similarity measures, face detection, speech detection. Linear classifiers, linear discriminant analysis

• Classification: Non-linear classifiers, neural nets, kernels, generative models, non-parametric methods. Real-world applications of classification models

• Clustering interlude. K-means, Gaussian Mixture Models, Expectation-Maximization algorithm

• Time series and dynamical models: Classification and similarity, time warping models, Markov models

• Mixed signals: Array processing, beamforming, independent component analysis, MIMO/SIMO models, under-constrained separation, spectral factorizations

• Matrix factorizations and bag-of-features models: Non-negative Matrix Factorization and Probabilistic Latent Semantic Decompositions, bag models, Convolutive decompositions

• Deep Learning. Boltzman Machines, Neural Networks, Convolutional and Recurrent models, etc.

• Missing data techniques and tracking

• Special Topic: Computer Vision

• Special Topic: Machine listening and Music Information Retrieval

• Special Topic: Speech

• Special Topic: Compressive Sensing

Lectures

Intro and Linear Algebra [PDF, Slides]

Probability and Stats [PDF, Slides]

Perception and Features [PDF, Slides]

Principal Component Analysis [PDF, Slides]

KPCA and Manifold Methods [PDF, Slides]

Detection and Matched Filters [PDF, Slides]

Decision theory & classifiers [PDF, Slides]

Nonlinear classifiers [PDF, Slides]

Classification bits and pieces [PDF, Slides]

Missing data & dynamical models [PDF, Slides]

Arrays & source separation [PDF, Slides]

Underconstrained separation [PDF, Slides]

Matrix Factorizations and beyond [PDF, Slides]

Compressive Sensing and Sparsity[PDF, Slides]

Privacy-Preserving Learning and Requests[PDF, Slides]

Problem Sets