Scribd Logo
Upload

Welcome to Scribd!

  • Lecture6 - Intro - To - Text - Speech - Image 1

    Uploaded by

    lubhawan parihar
    6 views23 pages
    This document discusses machine learning techniques for signal processing. It summarizes linear discriminant analysis (LDA) and how it finds a linear transformation that maximizes class separation by maximizing between-class distance and minimizing within-class covariance. LDA is compared to principal component analysis (PCA), where PCA focuses on variance while LDA focuses on discrimination. The document also discusses representations of data, including text representations using bag-of-words models and term frequency-inverse document frequency (TF-IDF), as well as audio representations using spectrograms.

    Original Description:

    Original Title

    Lecture6_Intro_to_text_speech_image 1

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    Report this Document
    This document discusses machine learning techniques for signal processing. It summarizes linear discriminant analysis (LDA) and how it finds a linear transformation that maximizes class separation by maximizing between-class distance and minimizing within-class covariance. LDA is compared to principal component analysis (PCA), where PCA focuses on variance while LDA focuses on discrimination. The document also discusses representations of data, including text representations using bag-of-words models and term frequency-inverse document frequency (TF-IDF), as well as audio representations using spectrograms.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Flag for inappropriate content
    6 views23 pages

    Lecture6 - Intro - To - Text - Speech - Image 1

    Uploaded by

    lubhawan parihar
    This document discusses machine learning techniques for signal processing. It summarizes linear discriminant analysis (LDA) and how it finds a linear transformation that maximizes class separation by maximizing between-class distance and minimizing within-class covariance. LDA is compared to principal component analysis (PCA), where PCA focuses on variance while LDA focuses on discrimination. The document also discusses representations of data, including text representations using bag-of-words models and term frequency-inverse document frequency (TF-IDF), as well as audio representations using spectrograms.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Flag for inappropriate content
    of 23

    MACHINE LEARNING

    FOR
    SIGNAL PROCESSING
    24-1-2022
    Sriram Ganapathy
    Electrical Engineering, Indian Institute of Science,
    sriramg@iisc.ac.in

    Debarpan Bhattacharya Hariom Choudhary Praneeth Linga Reddy


    EE, IISc, CSA, IISC ECE, IISC
    debarpanb@iisc.ac.in hariomc@iisc.ac.in praneethl@iisc.ac.in

    http://leap.ee.iisc.ac.in/sriram/teaching/MLSP22/

    Linear Discriminant Analysis


    (PRML 4.1.4)

    Without the Within Class Factor


    Linear Discriminant Analysis
    Find a linear transform with a criterion which
    maximizes the class separation

    • Maximize the between class distance in the projected space


    while minimizing the within class covariance

    ❖ Generalized Eigenvalue proble


    ❖ Eigenvectors of PRML - C. Bishop (Sec. 4.1.4, Sec. 4.1.6)

    Linear Discriminant Analysis

    Projecting on line joining means Fisher Discriminant

    PRML - C. Bishop (Sec. 4.1.4, Sec. 4.1.6)


    PCA versus LDA
    PCA versus LDA

    PCA
    LDA

    PRML - C. Bishop (Sec. 4.1.4, Sec. 4.1.6)


    Representations of data
    Knowledge based representations

    ❖ Text representation
    ❖ Audio representations
    s

    Text representations

    ❖ Bag of words mode


    ❖ Term frequency - inverse document frequency
    l

    TF-IDF
    Audio representations

    ❖ Spectrogram
    ❖ Computed using magnitude of Fourier transform
    s

    Spectrogram Example

    You might also like