Subject: Subject Sylbus: Nonlinear Signal Processing Using Geometric Methods - 048969 (Current)

Nonlinear Signal Processing Using Geometric Methods - 048969
Credit
Points
2.0
 
Given In
Semester
a
 
  Lecture Exercise Laboratory Project or
Seminar
House
Work
Weekly
Hours
2        

Determination of the grade according to progress during the semester.


Prerequisites: ( Intro. to Digital Signal Processing 044198
and Random Signals 044202 )


The Goal of This Course Is to Understand the Mathematical Foundation of Many Recent Methods for Interinsic Modeling and Geometry-Assisted Signal Processing. Novel Methods That Extend the Classical Fourier Analysis and Enable Natural Data-Driven Parameterization of Signals Without Prior Knowledge of Models. in Addition, Recent Nonlinear Filtering Methods Based on Data-Driven Geometric Models Will Be Presented. the Course Will Cover Topics from Harmonic Analysis, Graph Theory, Differential Geometry, Nonlinear Filtering, and Stochastic Diffusion Processes. Applications to Biomedical Signal Analysis, Audio and Speech Processing, and High Outcomes: at the End of the Course the Student Eill Be Able:
1. to Define Fundamental Terms, Including "Intrinsic Modeling"
2. to Analyze Special Cases of Signals and Systems Analytically, Build Intrinsic Metrics, and Prove Their Main Properties
3. to Implement (in Matlab) An Algorithm for Building Intrinsic Models of Synthetic and Real Signals.


Timetable to semester 01/2017 2017/2018 Winter Semester
RoomBuildingHourdayLecturerExercise
Lecture
no.Registering
Group
353בנ.מאייר14:30-16:30SundayAssistant Professor Talmon RonenLecture1010


Textbooks
PublishedPublisherAuthorsBook
2003neural computationbelkin, m. and niyogi, p.laplacian eigenmaps for dimensionality reduction and data representation
2008 singer, a. and coifman, r.rnon linear independent component analysis with diffusion maps
1997 chung f.rspctral graph theory

Created in 23/11/2017 Time 05:37:52