Income Group Classification Install Import - ConfusionMatrixDisplay Who=Read Data=Copy Head() tail() shape _ _ info() duplicated().sum() .isnull().sum() .describe().T
histogram_boxplot labeled_barplot stacked_barplot distribution_plot_wrt_target df=data.copy() Univariate Analysis //// Observations on workclass //// Observations on native_country //// Observations on salary Bivariate analysis //// Correlation check //// salary vs sex //// salary vs education //// salary vs occupation //// salary vs workclass //// salary vs age //// salary vs working_hours_per_week ((untitled summary plot for salary against several parameters/factors )) // Data Preprocessing ** Dropping capital_gain and capital_loss /// Outlier Detection /// Outlier Treatment treat_outliers treat_outliers_all numerical_col=data.select_dtypes() data=treat_outliers_all() plt.figure(figsize=(20, 30)) for i, variable in enumerate(numerical_col): plt ((parameters)) plt.show() /// Data Preparation for Modeling ** Encoding >50K as 0 and <=50K as 1 to find underprivileged data["salary"]=data["salary"].apply(lambda x: 1 if x == " <=50K" else 0) Creating training and test sets. X = data.drop(["salary"], axis=1) Y=data["salary"] X=sm.add_constant(X) Q/ What is sm? statsmodels.api A/ # To build model for prediction import statsmodels.stats.api as sms from statsmodels.stats.outliers_influence import variance_inflation_factor import statsmodels.api as sm from statsmodels.tools.tools import add_constant from sklearn.linear_model import LogisticRegression
X=pd.get_dummies(X,drop_first=True) X_train,X_tesst,y_train,y_tesst=train_test_split( X,Y,test_size=0.30) prints -> Training set and test set both have 53 cols
Model Building - Logistic Regression