def model_build(alg_name, params):
alg = None
if alg_name == 'SVM':
if params['kernel'] == 'linear':
alg = SVC(C=params['C'], probability=True, kernel='linear')
elif params['kernel'] == 'rbf':
alg = SVC(C=params['C'], gamma=params['gammas'], probability=True, kernel=params['kernel'])
elif params['kernel'] == 'poly':
alg = SVC(C=params['C'], degree=params['degree'], probability=True, kernel=params['kernel'])
elif alg_name == 'KNN':
alg = KNeighborsClassifier(n_neighbors=params['K'], weights=params['weights'], leaf_size=params['leaf_size'])
elif alg_name == 'Random Forest':
alg = RandomForestClassifier(n_estimators=params['n_estimators'], criterion=params['criterion'],
max_features=params['max_features'], random_state=1234)
elif alg_name == 'LightGBM':
alg = lgb.LGBMClassifier(learning_rate=params['learning_rate'], num_leaves=params['num_leaves'],
n_estimators=params['n_estimators'], objective=params['objective'])
elif alg_name == 'XGBoost':
alg = XGBClassifier(objective=params['objective'], eval_metrics=params['eval_metrics'],
learning_rate=params['learning_rate'], max_depth=params['max_depth'])
elif alg_name == 'Naive Bayes':
if params['distribution'] == 'Multinomial Naive Bayes':
alg = naive_bayes.MultinomialNB(alpha=params['alpha'], fit_prior=params['fit_prior'])
elif params['distribution'] == 'Gaussian Naive Bayes':
alg = naive_bayes.GaussianNB()
elif params['distribution'] == 'Complement Naive Bayes':
alg = naive_bayes.ComplementNB(alpha=params['alpha'], fit_prior=params['fit_prior'], norm=params['norm'])
elif params['distribution'] == 'Bernoulli Naive Bayes':
alg = naive_bayes.BernoulliNB(alpha=params['alpha'], fit_prior=params['fit_prior'],
binarize=params['binarize'])
elif params['distribution'] == 'Categorical Naive Bayes':
alg = naive_bayes.CategoricalNB(alpha=params['alpha'], fit_prior=params['fit_prior'])
return alg
X[:] = X[:] + abs(np.min(X[:]))
# print('After all values made negative, X = \n', X, '\n')
# split dataset into training and testing partitions
train_split = 0.7
train_all, test_all = split_partition(X, y, train_split)
# formally classify colours (not using Naive-Bayes)
Train = Partition(train_all, n_classes)
Test = Partition(test_all, n_classes)
# create Naive Bayes classifier model
train_all_X = train_all[:, :n_classes]
train_all_y = train_all[:, n_classes]
time_start = time.time()
nb_model = skl_nb.CategoricalNB().fit(train_all_X, train_all_y)
time_total = time.time() - time_start
# make predictions
test_all_X = test_all[:, :n_classes]
test_all_y = test_all[:, n_classes]
y_pred = nb_model.predict(test_all_X)
#%% merge NB-predicted y (classifications) and X (co-ordinates)
y_pred = np.reshape(y_pred, (y_pred.shape[0], 1))
Xy_pred = np.concatenate((test_all_X, y_pred), axis=1)
# create separate arrays for NB-classified co-ordinates
X_green_pred, X_blue_pred = split_colour(Xy_pred, n_classes)
# Performance Metrics:
def convert_to_numeric_values(df):
converted_df = df.copy()
converted_df = converted_df.replace({"history": {"bad": 0, "fair": 1, "excellent": 2},
"income": {"low": 0, "high": 1},
"term": {3: 0, 10: 1}})
return converted_df
loan_df = pd.read_csv("data/loans.csv")
numeric_loan_df = convert_to_numeric_values(loan_df)
print(numeric_loan_df)
feature_names = loan_df.columns.values[:-1]
X = numeric_loan_df[feature_names]
y = numeric_loan_df["risk"]
naive_bayes_model = sk_naive_bayes.CategoricalNB()
naive_bayes_model.fit(X, y)
X_probabilities = naive_bayes_model.predict_proba(X)[:, 1]
X_probabilities_log = naive_bayes_model.predict_log_proba(X)[:, 1]
loan_df["probability"] = X_probabilities
loan_df["log_probability"] = X_probabilities_log
print(loan_df)
labelEnc = preprocessing.LabelEncoder()
yTrain = labelEnc.fit_transform(subData['insurance'])
yLabel = labelEnc.inverse_transform([0, 1])
uGroup_Size = np.unique(subData['group_size'])
uHomeowner = np.unique(subData['homeowner'])
uMarried_Couple = np.unique(subData['married_couple'])
featureCategory = [uGroup_Size, uHomeowner, uMarried_Couple]
print(featureCategory)
featureEnc = preprocessing.OrdinalEncoder(categories=featureCategory)
xTrain = featureEnc.fit_transform(
subData[['group_size', 'homeowner', 'married_couple']])
_objNB = naive_bayes.CategoricalNB(alpha=1.0e-10)
thisModel = _objNB.fit(xTrain, yTrain)
print('Number of samples encountered for each class during fitting')
print(yLabel)
print(_objNB.class_count_)
print('\n')
print('Probability of each class:')
print(yLabel)
print(np.exp(_objNB.class_log_prior_))
print('\n')
#%%
print('---QUESTION 1B---')
RowWithColumn(rowVar=subData['insurance'],
data_train = pd.read_csv("regl_data/juegos_entrenamiento.txt",
sep=' ',
header=None)
data_test = pd.read_csv("regl_data/juegos_validacion.txt",
sep=' ',
header=None)
X = data_train.iloc[:, :2]
y = data_train[[2]]
X_test = data_test.iloc[:, :2]
y_test = data_test[[2]]
# entrenando el clasificador bayesiano ingenuo
cat = nb.CategoricalNB()
estimator = cat.fit(X, np.ravel(y))
print("params: ", estimator.get_params())
pred_train = estimator.predict(X_test)
suma = (np.ravel(y_test) != pred_train).sum()
prec = (100 - (suma / X.shape[0]) * 100)
#new_data = cat.trans
print(suma, prec)
# np.random.seed(0)
# W = np.random.uniform(0, 1, size=(X.shape[1], 1))
# def sigmoid(z):
# return (1 / 1 + np.exp(-z))