Scribd Logo
Upload

Welcome to Scribd!

  • 15 views

    A2 Vishal Borra

    Uploaded by

    vishal.borra
    The document defines functions for preprocessing data, applying clustering algorithms, and evaluating clustering performance. It loads clustering algorithms like MiniBatchKMeans, SpectralClustering, and OPTICS. For different datasets, it applies the algorithms, calculates the Davies-Bouldin Index metric, and identifies the best performing algorithm and number of clusters. It plots the clustered data and prints the results.

    Copyright:

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

    A2 Vishal Borra

    Uploaded by

    vishal.borra
    15 views2 pages
    The document defines functions for preprocessing data, applying clustering algorithms, and evaluating clustering performance. It loads clustering algorithms like MiniBatchKMeans, SpectralClustering, and OPTICS. For different datasets, it applies the algorithms, calculates the Davies-Bouldin Index metric, and identifies the best performing algorithm and number of clusters. It plots the clustered data and prints the results.

    Copyright

    © © All Rights Reserved

    Available Formats

    TXT, PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    The document defines functions for preprocessing data, applying clustering algorithms, and evaluating clustering performance. It loads clustering algorithms like MiniBatchKMeans, SpectralClustering, and OPTICS. For different datasets, it applies the algorithms, calculates the Davies-Bouldin Index metric, and identifies the best performing algorithm and number of clusters. It plots the clustered data and prints the results.

    Copyright:

    © All Rights Reserved
    Download as TXT, PDF, TXT or read online from Scribd
    Download as txt, pdf, or txt
    15 views2 pages

    A2 Vishal Borra

    Uploaded by

    vishal.borra
    The document defines functions for preprocessing data, applying clustering algorithms, and evaluating clustering performance. It loads clustering algorithms like MiniBatchKMeans, SpectralClustering, and OPTICS. For different datasets, it applies the algorithms, calculates the Davies-Bouldin Index metric, and identifies the best performing algorithm and number of clusters. It plots the clustered data and prints the results.

    Copyright:

    © All Rights Reserved
    Download as TXT, PDF, TXT or read online from Scribd
    Download as txt, pdf, or txt
    of 2

    import pandas as pd

    from sklearn.cluster import MiniBatchKMeans, SpectralClustering, OPTICS


    from sklearn.metrics import davies_bouldin_score
    import matplotlib.pyplot as plt

    # Data preprocessing function


    def preprocess_data(file_path):
    # Reading the data from CSV
    data = pd.read_csv(file_path, header=None)
    # Normalizing the data
    data_normalized = (data - data.mean()) / data.std()
    return data_normalized

    # Function to apply clustering and calculate the Davies-Bouldin Index


    def apply_clustering_and_evaluate(data, algorithm):
    # Applying the clustering algorithm to the data
    labels = algorithm.fit_predict(data)
    # Calculating Davies-Bouldin Index
    db_index = davies_bouldin_score(data, labels)
    return labels, db_index

    # Defin ing the clustering algorithms


    algorithms = {
    'MiniBatchKMeans': MiniBatchKMeans(n_clusters=3, batch_size=100),
    'SpectralClustering': SpectralClustering(n_clusters=3,
    affinity='nearest_neighbors'),
    'OPTICS': OPTICS(min_samples=5, xi=0.05, min_cluster_size=0.1)
    }

    # Main execution loop


    for path in [
    (r"C:\Users\Vishal\Desktop\DM2\D01.csv"),
    (r"C:\Users\Vishal\Desktop\DM2\D02.csv"),
    (r"C:\Users\Vishal\Desktop\DM2\D03.csv")
    ]:
    data = preprocess_data(path)

    best_davies_bouldin_score = float('inf')
    best_algorithm_name = None
    best_labels = None

    for name, alg in algorithms.items():


    labels, db_index = apply_clustering_and_evaluate(data, alg)
    if db_index < best_davies_bouldin_score:
    best_davies_bouldin_score = db_index
    best_algorithm_name = name
    best_labels = labels

    # Counting number of clusters


    num_clusters = len(set(best_labels))

    # Plotting the results


    plt.scatter(data[0], data[1], c=best_labels, cmap='viridis', marker='o')
    plt.title(f'{best_algorithm_name} - Davies-Bouldin Index:
    {best_davies_bouldin_score:.2f}, Number of Clusters: {num_clusters}')
    plt.xlabel('Feature 1')
    plt.ylabel('Feature 2')
    plt.show()
    print(f"The best algorithm for {path} is {best_algorithm_name} with a Davies-
    Bouldin Index of {best_davies_bouldin_score:.2f} and {num_clusters} clusters")

    You might also like