• ID3 Algorithm
  • Mining Big Data
  • K Nearest Neighbors Algorithm KNN




    Download 474,75 Kb.
    Pdf ko'rish
    bet5/8
    Sana20.05.2024
    Hajmi474,75 Kb.
    #245707
    1   2   3   4   5   6   7   8
    Bog'liq
    2.1 ga oid

    K Nearest Neighbors Algorithm KNN
    The k-nearest neighbors (KNN) algorithm is a simple, 
    easy-to-implement supervised machine learning algorithm that 
    can be used to solve both classification and regression problems. 
    KNN is a non-parametric, simple learning algorithm. Its pur-
    pose is to use a database in which the data points are separated 
    into several classes to predict the classification of a new sample 
    point.
    Naive Bayes
    The Naive Bayes algorithm is based on the Bayesian 
    theorem. It is particularly used when the dimensionality of the 
    inputs is high. The Bayesian algorithm is capable of calculating 
    the possible output. That is based on the input. Naive Bayes can 
    often outperform more sophisticated classification methods.
    ID3 Algorithm
    This Data Mining Algorithm starts with the main set as 
    the root hub. On every cycle, it emphasizes through every un-
    used attribute of the set and figures. That the entropy of attribute. 
    At that point chooses the attribute. That has the smallest entropy 
    value. The set is S then split by the selected attribute to produce 
    subsets of the information. This Data Mining algorithms proceed 
    to recurse on each item in a subset. Also, considering only items 
    never selected before.
    Mining Big Data
    In recent years, massive amounts of data are generated 
    every moment in different fields such as Internet, bank, health 
    care, social media and physical systems, this is known as big 
    data. Valuable information can be extracted from this big data 
    by using data mining. Traditional data mining techniques can 
    find out potential useful information, valuable relationships and 
    patterns in the data. The extracted information supports decision 


    J Comput Sci Software Dev 2022 | Vol 2: 303
    JScholar Publishers
    4
    making process and makes some predictions [16,17]. Multiple 
    applications gain benefits from data mining such as education, 
    science, health and smart cities [13]. The traditional data min-
    ing techniques unsuitable to deal with big data due to limitations 
    of these techniques in dealings with characteristics of big data 
    [16,17]. Big data mining is the ability to extracting valuable and 
    beneficial information from huge datasets that is due to its het-
    erogeneity, volume and velocity, it was not possible to do it [13]. 
    New requirements and techniques need to manage and min-
    ing big data, in order to fulfil these requirements MapReduce 
    and Hadoop introduced [16].

    Download 474,75 Kb.
    1   2   3   4   5   6   7   8




    Download 474,75 Kb.
    Pdf ko'rish