Scribd Logo
Upload

Welcome to Scribd!

    • 54 views

    SVM - An Essay

    Uploaded by

    Siladitya Dash
    Supervised machine learning involves feeding a machine both feature vectors and labels to build a model that can label future vectors. One such algorithm is support vector machines (SVM), which can differentiate entities like cars and bikes. SVM works by plotting each data item in an n-dimensional space based on feature values, and finding a hyperplane that best separates the classes with the maximum margin between them. The hyperplane is used to classify new data points based on which side they fall within.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd

    SVM - An Essay

    Uploaded by

    Siladitya Dash
    54 views1 page
    Supervised machine learning involves feeding a machine both feature vectors and labels to build a model that can label future vectors. One such algorithm is support vector machines (SVM), which can differentiate entities like cars and bikes. SVM works by plotting each data item in an n-dimensional space based on feature values, and finding a hyperplane that best separates the classes with the maximum margin between them. The hyperplane is used to classify new data points based on which side they fall within.

    Original Description:

    ML research

    Original Title

    1828033_SVM - An Essay

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    Supervised machine learning involves feeding a machine both feature vectors and labels to build a model that can label future vectors. One such algorithm is support vector machines (SVM), which can differentiate entities like cars and bikes. SVM works by plotting each data item in an n-dimensional space based on feature values, and finding a hyperplane that best separates the classes with the maximum margin between them. The hyperplane is used to classify new data points based on which side they fall within.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    54 views1 page

    SVM - An Essay

    Uploaded by

    Siladitya Dash
    Supervised machine learning involves feeding a machine both feature vectors and labels to build a model that can label future vectors. One such algorithm is support vector machines (SVM), which can differentiate entities like cars and bikes. SVM works by plotting each data item in an n-dimensional space based on feature values, and finding a hyperplane that best separates the classes with the maximum margin between them. The hyperplane is used to classify new data points based on which side they fall within.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    You are on page 1of 1

    WHAT I KNOW ABOUT SUPERVISED LEARNING USING SVM ALGORITHM

    There are three basic types of machine learning. They are Supervised, Unsupervised and
    Semi-supervised machine learning.

    In case of supervised machine learning, the machine is fed with both feature vectors and labels and it
    uses certain algorithms to come up with a model which can label any future vectors that are given to
    it. In simple words, we train the machine to come up with solutions by providing examples and their
    solutions.

    Suppose we have to differentiate whether the entity which has been input is a car or bike. For that,
    we have to train the machine by providing a data sheet which contains feature vectors, basically,
    which can help the machine differentiate between a car and a bike. For example, if the entity input
    has two wheels, or it is small then we can label as a bike, by using a number 0, for instance. And if it is
    big in size, or has four wheels we can label it as 1,i.e., a car.

    Since the inputs are machine readable but the outputs are still human readable only (car, bike, etc.),
    we need to convert them likewise. SVM is one such algorithm which can be used to get the job done.

    In this algorithm, we plot each data item as a point in n-dimensional space (where n is number of
    features we have) with the value of each feature vector being the value of a particular coordinate.
    Then, we perform classification by finding the hyper-plane that differentiate the two classes. A hyper
    plane is selected in such a way that it segregates the different classes in the best possible way. For
    example, if we have two classes - cars and bikes - then the hyper plane has to be selected in such a
    way that all car entities are above the line (for example) and all bike entities are below the line and
    they are as far away from the line as possible (for greater accuracy).

    Suppose the hyper plane ( a straight line for 2 dimensional studies) has the equation y = wx + b = 0. It
    is called linear SVM, in this case. The sum of the distances of the closest entities of both the classes is
    known as the margin. Let the line touching the closest entity of the upper class (say for ‘bikes’) has
    the the equation wx + b = 1 and likewise, wx + b = -1 for the lower class (which is for ‘cars’. w and b
    are called weights and bias respectively. They are compared with the feature vectors and determine
    the label value. Then any entity which we input in the testing period that has label value (y) greater
    than 1 will be classified as bike (let’s say) and below -1 will be classified as car. And the farther the
    prediction from the hyper plane, the more accurate it is. This is how a machine can be trained to
    identify the difference between entities by using SVM algorithm.

    SILADITYA DASH

    1828033

    2nd YEAR, CSSE

    You might also like