Skip to content
The Chat Archive
o1-preview-2024-09-12· OpenAI WildChat-4.8M (AI2)

import pandas as pd import numpy as np from sklearn.preprocessing import StandardScaler…

  1. User

    import pandas as pd import numpy as np from sklearn.preprocessing import StandardScaler from sklearn.model_selection import train_test_split, KFold from sklearn.metrics import accuracy_score from sklearn.neighbors import NearestNeighbors

    class BMFK: def init(self, n_neighbors=5, m=2, p=2, q=2): self.n_neighbors = n_neighbors self.m = m # fuzzy strength parameter self.p = p # Bonferroni p parameter self.q = q # Bonferroni q parameter

    def fit(self, X, y):
        self.X = X
        self.y = y
        self.classes = np.unique(y)
        self.nn = NearestNeighbors(n_neighbors=self.n_neighbors, metric='minkowski', p=2)
        self.nn.fit(X)
    
    def bonferroni_mean(self, values):
        n = len(values)
        if n <= 1:
            return np.mean(values)  # Return simple mean if there's only one value
        sum_pq = 0
        for i in range(n):
            for j in range(n):
                if i != j:
                    sum_pq += values[i]**self.p * values[j]**self.q
        return (sum_pq / (n * (n-1)))**(1 / (self.p + self.q))
    
    def predict(self, X):
        predictions = []
        for x in X:
            distances, indices = self.nn.kneighbors([x])
            neighbors = self.X[indices[0]]
            neighbor_labels = self.y[indices[0]]
            
            # Calculate fuzzy memberships
            memberships = 1 / (distances[0] ** (2 / (self.m - 1)) + 1e-8)
            memberships /= np.sum(memberships)
            
            # Calculate class memberships using Bonferroni mean
            class_memberships = {}
            for c in self.classes:
                class_indices = neighbor_labels == c
                if np.any(class_indices):
                    class_memberships[c] = self.bonferroni_mean(memberships[class_indices])
                else:
                    class_memberships[c] = 0
            
            # Predict the class with highest membership
            predictions.append(max(class_memberships, key=class_memberships.get))
        
        return np.array(predictions)
    

    Load the dataset

    file_path = "PCOS_data_without_infertility.xlsx" df = pd.read_excel(file_path, sheet_name="Full_new")

    Data Preprocessing

    df = df.drop(columns=['Sl. No', 'Patient File No.', 'Unnamed: 44']) df = df.apply(pd.to_numeric, errors='coerce') df.fillna(df.median(), inplace=True)

    Encode categorical columns

    categorical_columns = ['Blood Group', 'Cycle(R/I)', 'Pregnant(Y/N)', 'Weight gain(Y/N)', 'hair growth(Y/N)', 'Skin darkening (Y/N)', 'Hair loss(Y/N)', 'Pimples(Y/N)', 'Fast food (Y/N)', 'Reg.Exercise(Y/N)'] for col in categorical_columns: if col in df.columns: df[col] = df[col].astype('category').cat.codes

    Split into features and target

    X = df.drop(columns=['PCOS (Y/N)']) y = df['PCOS (Y/N)'].values # Convert to numpy array

    Standardize features

    scaler = StandardScaler() X_scaled = scaler.fit_transform(X)

    Define the different feature sets

    ensemble_selected_features = ['Follicle No. (L)', 'hair growth(Y/N)', 'Follicle No. (R)', 'Cycle(R/I)', 'Fast food (Y/N)', 'AMH(ng/mL)', 'Skin darkening (Y/N)', 'Weight gain(Y/N)', 'Pimples(Y/N)', 'Cycle length(days)', 'Hip(inch)', 'Weight (Kg)', 'FSH/LH', 'FSH(mIU/mL)']

    rl_selected_features = [' Age (yrs)', 'Weight (Kg)', 'Pulse rate(bpm) ', 'RR (breaths/min)', 'Hb(g/dl)', 'Cycle(R/I)', 'Cycle length(days)', 'Marraige Status (Yrs)', ' I beta-HCG(mIU/mL)', 'FSH(mIU/mL)', 'Waist(inch)', 'Waist:Hip Ratio', 'TSH (mIU/L)', 'AMH(ng/mL)', 'Vit D3 (ng/mL)', 'PRG(ng/mL)', 'RBS(mg/dl)', 'Weight gain(Y/N)', 'hair growth(Y/N)', 'Skin darkening (Y/N)', 'Pimples(Y/N)', 'BP _Systolic (mmHg)', 'BP _Diastolic (mmHg)', 'Follicle No. (L)', 'Follicle No. (R)']

    proposed_selected_features = ['Follicle No. (L)', 'hair growth(Y/N)', 'Follicle No. (R)', 'Cycle(R/I)', 'Fast food (Y/N)', 'Skin darkening (Y/N)', 'Cycle length(days)', 'FSH/LH']

    1. BMFK on Ensemble Selected Features

    X_ensemble = X_scaled[:, [X.columns.get_loc(feature) for feature in ensemble_selected_features]]

    2. BMFK on RL Selected Features

    X_rl = X_scaled[:, [X.columns.get_loc(feature) for feature in rl_selected_features]]

    3. BMFK on Proposed BMFK Selected Features

    X_proposed = X_scaled[:, [X.columns.get_loc(feature) for feature in proposed_selected_features]]

    4. BMFK on All Features (no feature selection)

    X_all = X_scaled # Use all features

    Helper function to run cross-validation

    def run_bmfk_cv(X, y): kf = KFold(n_splits=10, shuffle=True, random_state=42) accuracies = [] for train_idx, test_idx in kf.split(X): X_train, X_test = X[train_idx], X[test_idx] y_train, y_test = y[train_idx], y[test_idx]

        model = BMFK(n_neighbors=5, m=2, p=1, q=1)
        model.fit(X_train, y_train)
        y_pred = model.predict(X_test)
        accuracy = accuracy_score(y_test, y_pred)
        accuracies.append(accuracy)
    
    return np.mean(accuracies)
    

    Run BMFK for all four cases

    accuracy_ensemble = run_bmfk_cv(X_ensemble, y) accuracy_rl = run_bmfk_cv(X_rl, y) accuracy_proposed = run_bmfk_cv(X_proposed, y) accuracy_all = run_bmfk_cv(X_all, y)

    print(f"BMFK on Ensemble Selected Features Accuracy: {accuracy_ensemble:.4f}") print(f"BMFK on RL Selected Features Accuracy: {accuracy_rl:.4f}") print(f"BMFK on Proposed Selected Features Accuracy: {accuracy_proposed:.4f}") print(f"BMFK on All Features Accuracy: {accuracy_all:.4f}")

    use svm here instead of bmfk

  2. Assistant
    import pandas as pd
    import numpy as np
    from sklearn.preprocessing import StandardScaler
    from sklearn.model_selection import KFold
    from sklearn.metrics import accuracy_score
    from sklearn.svm import SVC
    
    # Load the dataset
    file_path = "PCOS_data_without_infertility.xlsx"
    df = pd.read_excel(file_path, sheet_name="Full_new")
    
    # Data Preprocessing
    df = df.drop(columns=['Sl. No', 'Patient File No.', 'Unnamed: 44'])
    df = df.apply(pd.to_numeric, errors='coerce')
    df.fillna(df.median(), inplace=True)
    
    # Encode categorical columns
    categorical_columns = ['Blood Group', 'Cycle(R/I)', 'Pregnant(Y/N)', 
                           'Weight gain(Y/N)', 'hair growth(Y/N)', 
                           'Skin darkening (Y/N)', 'Hair loss(Y/N)', 
                           'Pimples(Y/N)', 'Fast food (Y/N)', 
                           'Reg.Exercise(Y/N)']
    for col in categorical_columns:
        if col in df.columns:
            df[col] = df[col].astype('category').cat.codes
    
    # Split into features and target
    X = df.drop(columns=['PCOS (Y/N)'])
    y = df['PCOS (Y/N)'].values  # Convert to numpy array
    
    # Standardize features
    scaler = StandardScaler()
    X_scaled = scaler.fit_transform(X)
    
    # Define the different feature sets
    ensemble_selected_features = ['Follicle No. (L)', 'hair growth(Y/N)', 'Follicle No. (R)', 
                                  'Cycle(R/I)', 'Fast food (Y/N)', 'AMH(ng/mL)', 'Skin darkening (Y/N)', 
                                  'Weight gain(Y/N)', 'Pimples(Y/N)', 'Cycle length(days)', 
                                  'Hip(inch)', 'Weight (Kg)', 'FSH/LH', 'FSH(mIU/mL)']
    
    rl_selected_features = [' Age (yrs)', 'Weight (Kg)', 'Pulse rate(bpm) ', 'RR (breaths/min)', 'Hb(g/dl)', 
                            'Cycle(R/I)', 'Cycle length(days)', 'Marraige Status (Yrs)', '  I   beta-HCG(mIU/mL)', 
                            'FSH(mIU/mL)', 'Waist(inch)', 'Waist:Hip Ratio', 'TSH (mIU/L)', 'AMH(ng/mL)', 
                            'Vit D3 (ng/mL)', 'PRG(ng/mL)', 'RBS(mg/dl)', 'Weight gain(Y/N)', 'hair growth(Y/N)', 
                            'Skin darkening (Y/N)', 'Pimples(Y/N)', 'BP _Systolic (mmHg)', 'BP _Diastolic (mmHg)', 
                            'Follicle No. (L)', 'Follicle No. (R)']
    
    proposed_selected_features = ['Follicle No. (L)', 'hair growth(Y/N)', 'Follicle No. (R)', 
                                  'Cycle(R/I)', 'Fast food (Y/N)', 'Skin darkening (Y/N)', 
                                  'Cycle length(days)', 'FSH/LH']
    
    # 1. SVM on Ensemble Selected Features
    X_ensemble = X_scaled[:, [X.columns.get_loc(feature) for feature in ensemble_selected_features]]
    
    # 2. SVM on RL Selected Features
    X_rl = X_scaled[:, [X.columns.get_loc(feature) for feature in rl_selected_features]]
    
    # 3. SVM on Proposed Selected Features
    X_proposed = X_scaled[:, [X.columns.get_loc(feature) for feature in proposed_selected_features]]
    
    # 4. SVM on All Features (no feature selection)
    X_all = X_scaled  # Use all features
    
    # Helper function to run cross-validation
    def run_svm_cv(X, y, svm_params=None):
        if svm_params is None:
            svm_params = {}
        kf = KFold(n_splits=10, shuffle=True, random_state=42)
        accuracies = []
        for train_idx, test_idx in kf.split(X):
            X_train, X_test = X[train_idx], X[test_idx]
            y_train, y_test = y[train_idx], y[test_idx]
            
            model = SVC(**svm_params)
            model.fit(X_train, y_train)
            y_pred = model.predict(X_test)
            accuracy = accuracy_score(y_test, y_pred)
            accuracies.append(accuracy)
        
        return np.mean(accuracies)
    
    # Run SVM for all four cases
    svm_params = {'kernel': 'rbf', 'C': 1, 'gamma': 'scale'}  # You can adjust these parameters
    
    accuracy_ensemble = run_svm_cv(X_ensemble, y, svm_params)
    accuracy_rl = run_svm_cv(X_rl, y, svm_params)
    accuracy_proposed = run_svm_cv(X_proposed, y, svm_params)
    accuracy_all = run_svm_cv(X_all, y, svm_params)
    
    print(f"SVM on Ensemble Selected Features Accuracy: {accuracy_ensemble:.4f}")
    print(f"SVM on RL Selected Features Accuracy: {accuracy_rl:.4f}")
    print(f"SVM on Proposed Selected Features Accuracy: {accuracy_proposed:.4f}")
    print(f"SVM on All Features Accuracy: {accuracy_all:.4f}")
    

    This code replaces the BMFK classifier with an SVM classifier and retains the same preprocessing and feature selection steps. It applies 10-fold cross-validation for each of the four feature sets using the SVM classifier and reports the mean accuracy for each. Adjust the svm_params dictionary to tune the SVM's hyperparameters as needed.