MachineLearning

MachineLearning.py

import pandas as pd
import numpy as np

from sklearn.model_selection import KFold
from sklearn.model_selection import cross_val_score
from sklearn.metrics import accuracy_score, confusion_matrix, classification_report
from sklearn.model_selection import train_test_split

# Modelos
from sklearn.naive_bayes import GaussianNB
from sklearn.naive_bayes import MultinomialNB
from sklearn.naive_bayes import ComplementNB
from sklearn.naive_bayes import BernoulliNB
from sklearn.naive_bayes import CategoricalNB

# FUNÇÃO GENÉRICA PARA OS MODELOS DE NAYVE BAYES
def f_modelo(funcao, funcao_name):
    # ==============
    # NAYVE BAYES
    # ==============
    print("============================ " + funcao_name)
    funcao.fit(x_treino, y_treino)
    prev_funcao = funcao.predict(x_teste)

    # print(confusion_matrix(y_teste, prev_funcao))
    # print(classification_report(y_teste, prev_funcao))

    print("Acurácia...: %.2f%%" % (accuracy_score(y_teste, prev_funcao) * 100.0))

    # CROSS-VALIDATION
    funcao_cv = funcao
    resultado_funcao_cv = cross_val_score(funcao_cv, previsores, alvo, cv=kfold)
    print("Acurácia CV: %.2f%%" % (resultado_funcao_cv.mean() * 100.0))
    print("\n")


# kfold = KFold(n_splits=30, shuffle=True, random_state=10)
kfold = KFold(n_splits=30, shuffle=True, random_state=0)

# CONFIGURAÇÕES DE IMPRESSÃO DE DADOS (APRESENTA TODAS AS COLUNAS DO DATASET NO CONSOLE, POR PADRÃO ERA APENAS 5)
desired_width = 320
pd.set_option('display.width', desired_width)
np.set_printoptions(linewidth=desired_width)
pd.set_option('display.max_columns', 15)

# CARREGA O CSV DIRETAMENTE DA PASTA DO PROJETO
ds_stroke = pd.read_csv('STROKE.csv', sep=',', encoding='iso-8859-1')

"""
    print(ds_stroke.head(20))
    print(ds_stroke.dtypes)
    print(ds_stroke.shape)
"""

# VERIFICA VALORES NULOS
# TAG BMI -> MUDA NULOS PARA 0
ds_stroke['bmi'].fillna(0, inplace=True)

# PADRONIZAÇÃO DOS DADOS
"""
    Dados que serão padronizados
    1   gender             5110 non-null   object
    5   ever_married       5110 non-null   object 
    6   work_type          5110 non-null   object 
    7   Residence_type     5110 non-null   object
    10  smoking_status     5110 non-null   object
"""
# GENDER
ds_stroke['gender'].replace({'Female': 0, 'Male': 1, 'Other': 2}, inplace=True)

# EVER_MARRIED
ds_stroke['ever_married'].replace({'Yes': 1, 'No': 0}, inplace=True)

# WORK_TYPE
ds_stroke['work_type'].replace({'Private': 0, 'Self-employed': 1, 'children': 2, 'Govt_job': 3, 'Never_worked': 4},
                               inplace=True)

# RESIDENCE_TYPE
ds_stroke['Residence_type'].replace({'Urban': 0, 'Rural': 1}, inplace=True)

# SMOKING_STATUS
ds_stroke['smoking_status'].replace({'never smoked': 0, 'Unknown': 1, 'formerly smoked': 2, 'smokes': 3}, inplace=True)

# PREVISORES / ALVO
previsores = ds_stroke.iloc[:, 0:11].values
alvo = ds_stroke.iloc[:, 11].values

# BASE DE TESTES / TREINO
x_treino, x_teste, y_treino, y_teste = train_test_split(previsores, alvo, test_size=0.20, random_state=0)

# CHAMA OS MÉTODOS DE APRENDIZAGEM
f_modelo(GaussianNB(), "Gauss")

f_modelo(MultinomialNB(), "Multinomial")

f_modelo(ComplementNB(), "Complement")

f_modelo(BernoulliNB(), "Bernoulli")

# f_modelo(CategoricalNB(), "Categorical") deu erro :)