Data Preprocessing
from sklearn.preprocessing import <classname>
StandardScaler, MinMaxScaler, RobustScaler
QuantileTransformer , PowerTransformer, FunctionTransformer
KBinsDiscretizer , PolynomialFeatures , Normalizer
scaler = StandardScaler()
# Apply a user-defined function to the data
transformer = FunctionTransformer(np.log1p)
# Discretize features into k bins
discretizer = KBinsDiscretizer(n_bins=3, encode='ordinal', strategy='uniform')
poly_features = PolynomialFeatures(degree=2)
X_scaled = <object>.fit_transform(X)
|
Encoding Categorical Data
from sklearn.preprocessing import <classname>
LabelEncoder , OneHotEncoder , OrdinalEncoder , LabelBinarizer
tb = OneHotEncoder()
le = LabelEncoder()
lb = LabelBinarizer()
y = le.fit_transform(['Yes', 'No', 'Yes'])
y = lb.fit_transform(['Yes', 'No', 'Yes'])
X_encoded = tb.fit_transform(X)
|
Handling missing values
from sklearn.impute import SimpleImputer, KNNImputer , IterativeImputer , MissingIndicator
from sklearn.experimental import enable_iterative_imputer
imputer = SimpleImputer(strategy='mean')
imputer = KNNImputer(n_neighbors=2)
imputer = IterativeImputer(random_state=0)
indicator = MissingIndicator()
X_imputed = imputer.fit_transform(X)
|
Feature Selection:
from sklearn.feature_selection import
SelectKBest ,SelectPercentile, SelectFromModel, VarianceThreshold, RFE, RFECV,
SequentialFeatureSelector
|
Dimensionality Reduction
from sklearn.decomposition import
PCA, IncrementalPCA, TruncatedSVD, KernelPCA, NMF, FastICA, LatentDirichletAllocation
pca = PCA(n_components=2)
kpca = KernelPCA(n_components=2, kernel='rbf')
tsne = TSNE(n_components=2)
X_new = any.fit_transform(X)
|
Pipelines:
from sklearn.pipeline import
Pipeline
FeatureUnion
ColumnTransformer
|
Supervised Learning Models:
|
Linear Models: LinearRegression, Ridge, Lasso, ElasticNet, LogisticRegression, SGDClassifier, SGDRegressor, Perceptron
|
|
Naive Bayes: GaussianNB, BernoulliNB, MultinomialNB,
|
|
Tree-Based Models: DecisionTreeClassifier, DecisionTreeRegresso,
|
|
Support Vector Machines (SVM): SVC, SVR, LinearSVC, LinearSVR, NuSVC, NuSVR, OneClassSVM
|
|
Nearest Neighbors: KNeighborsClassifier, KNeighborsRegressor, RadiusNeighborsClassifier, RadiusNeighborsRegressor
|
|
Neural Networks: MLPClassifier, MLPRegressor
|
|
Ensemble RandomForestClassifier, RandomForestRegressor, GradientBoostingClassifier, GradientBoostingRegressor, ExtraTreesClassifier, ExtraTreesRegressor, AdaBoostClassifier, AdaBoostRegressor
|
|
xgboost XGBClassifier, XGBRegressor
|
|
lightgbm LGBMClassifier, LGBMRegressor
|
|
catboost CatBoostClassifier, CatBoostRegressor,
|
from sklearn.linear_model import ,from sklearn.naive_bayes , from sklearn.tree
from sklearn.ensemble , from xgboost import , from lightgbm import
from catboost import , from sklearn.svm
from sklearn.neighbors
from sklearn.neural_network import
Semi-Supervised Learning:
LabelPropagation
LabelSpreading
|
Unsupervised Learning Models
|
Clustering: KMeans, AgglomerativeClustering ,DBSCAN, Birch, SpectralClustering
|
|
Dimensionality Reduction: PCA, IncrementalPCA, TruncatedSVD, KernelPCA, NMF, FastICA, LatentDirichletAllocation
|
Clustering
KMeans
AgglomerativeClustering
DBSCAN
Birch
SpectralClustering
|
Model Evaluation Metrics
|
Regression Metrics: mean_squared_error, r2_score, mean_absolute_error, explained_variance_score, median_absolute_error, mean_squared_log_error
|
|
Classification Metrics: accuracy_score, precision_score, recall_score, f1_score, roc_auc_score, average_precision_score, log_loss, confusion_matrix, classification_report
|
from sklearn.metrics import
|
|
|
