Scribd Logo
Upload

Welcome to Scribd!

  • 429 views

    Lazy Learners PDF

    Uploaded by

    Komal Modi
    This document discusses lazy learners, a type of machine learning algorithm. Lazy learners, such as k-nearest neighbors and case-based reasoning, store all training data without analyzing it and only perform analysis when classifying new data points. This allows lazy learners to have a fast training process but a slower classification process compared to eager learners. The document then describes the k-nearest neighbors algorithm in more detail and how it works to classify new data based on similarity to stored training examples.

    Copyright:

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

    Lazy Learners PDF

    Uploaded by

    Komal Modi
    429 views15 pages
    This document discusses lazy learners, a type of machine learning algorithm. Lazy learners, such as k-nearest neighbors and case-based reasoning, store all training data without analyzing it and only perform analysis when classifying new data points. This allows lazy learners to have a fast training process but a slower classification process compared to eager learners. The document then describes the k-nearest neighbors algorithm in more detail and how it works to classify new data based on similarity to stored training examples.

    Original Title

    Lazy Learners.pdf

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    This document discusses lazy learners, a type of machine learning algorithm. Lazy learners, such as k-nearest neighbors and case-based reasoning, store all training data without analyzing it and only perform analysis when classifying new data points. This allows lazy learners to have a fast training process but a slower classification process compared to eager learners. The document then describes the k-nearest neighbors algorithm in more detail and how it works to classify new data based on similarity to stored training examples.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    429 views15 pages

    Lazy Learners PDF

    Uploaded by

    Komal Modi
    This document discusses lazy learners, a type of machine learning algorithm. Lazy learners, such as k-nearest neighbors and case-based reasoning, store all training data without analyzing it and only perform analysis when classifying new data points. This allows lazy learners to have a fast training process but a slower classification process compared to eager learners. The document then describes the k-nearest neighbors algorithm in more detail and how it works to classify new data based on similarity to stored training examples.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    of 15

    Lazy Learners

    Vaibhav (16103472)
    Komal (16103474)
    Raj (16103475)
    Classification
    ● Data Mining technique based on machine learning.
    ● Used to classify each item in a set of data into one of a
    predefined set of classes or groups.
    ● More easy way to understand this :

    Assigning an object to a certain class based on its similarity to


    previous examples of other objects.

    ● Lazy Learners and Eager learners


    Lazy Learners
    ● Just store Data set without learning from it.
    ● Start classifying data when it receive Test data.
    ● So it takes less time learning and more time classifying data.
    ● Example : k-nearest neighbor, case based reasoning
    K-Nearest Neighbour
    ● This algorithm is typically used when we don’t have any prior
    knowledge about the distribution of the data
    ● Nearest neighbor algorithms are based on learning by analogy,
    that is, by comparing a given test tuple with training tuples
    similar to it
    Approach
    ● k-Nearest Neighbor stores all instances correspond to training data
    points in n-dimensional space
    ● An object is classified by a majority votes for its neighbour classes
    ● The object is then assigned the most common class among its K
    neighbours which is determined using a distance function
    Algorithm

    ● Calculate the distance between a new example (E) and all


    examples in the training set
    ● The Euclidean distance between 2 examples:
    An example: Suppose we want to test whether a special tissue
    paper is good or bad. We have 2 attributes: Acid durability and
    strength. We get a test sample with X1=3 and X2=7
    We calculate the Euclidean distance of the input with the training
    samples that we already have
    For k=3 we sort the distances and determine the nearest
    neighbours
    We gather the Y coordinate of the nearest neighbours. Applying
    simple majority to determine the category for the input
    Applications

    This algorithm is typically used for pattern detection

    Some applications include credit ratings, text classification,


    predicting voting patterns
    Case-Based Reasoning

    ● Case-based reasoning is the process of solving new problems


    based on the solutions of similar past problems.
    ● When a test tuple is given to classify it, a case-based reasoner
    will first check if an identical training case exists. If it is found,
    then the accompanying solution is returned.
    ● If no identical case is found, then the case-based reasoner will
    check for the training tuples which have identical components
    that are similar to the test tuple.
    Advantages

    ● The main advantage gained in employing a lazy learning


    method is that the target function will be approximated locally,
    such as in the k-nearest neighbor algorithm.
    ● lazy learning systems can simultaneously solve multiple
    problems and deal successfully with changes in the problem
    domain.
    Disadvantages

    ● The large space requirement to store the entire training


    dataset.
    ● Lazy learning methods are usually slower to evaluate.
    ● Particularly noisy training data increases the case base
    unnecessarily, because no abstraction is made during the
    training phase.
    ● Larger training data also entail increased cost.
    Thank You

    You might also like