Linear models
- Linear Regression: Modelo lineal para regresión
- How to split the data between training and test
- Ordinary Least Squares
- Analyzing the results of the model
- Logistic Regression: Modelo lineal para clasificación
- Sigmoid function
- Regulated linear models (Penalized regression)
- LASSO Regression (L1 Regularization)
- Ridge Regression (L2 Regularization)
- ElasticNet Regression (L1 & L2 Regularization)
- LARS: Least-Angle Regression
- SGD: Otra forma de modelos lineales
- Perceptron
- Vowpal Wabbit
- Variantes No Lineales
- Regresión Polinómica
- Generalized Additive Models (GAM)
Sowftware
Modelos lineales
| Sklearn (CPU) | RAPIDS (GPU) | |
|---|---|---|
| Linear Regression | sklearn.linear_model.LinearRegression |
cuml.LinearRegression |
| Logistic Regression | sklearn.linear_model.LogisticRegression |
cuml.LogisticRegression |
| Ridge Regression | sklearn.linear_model.Ridge |
cuml.Ridge |
| Lasso Regression | sklearn.linear_model.Lasso |
cuml.Lasso |
| ElasticNet Regression | sklearn.linear_model.ElasticNet |
cuml.ElasticNet |
| MiniBatch SGD Classifier | sklearn.linear_model.SGDClassifier |
cuml.MBSGDClassifier |
| MiniBatch SGD Regressor | sklearn.linear_model.SGDRegressor |
cuml.MBSGDRegressor |
| Mutinomial Naive Bayes | cuml.MultinomialNB |
|
| Stochastic Gradient Descent | cuml.SGD |
|
| Coordinate Descent | cuml.CD |
|
| Quasi-Newton | cuml.QN |
|
| Support Vector Machine Classifier | cuml.svm.SVC(kernel="linear") |
|
| Support Vector Machine Regressor | cuml.svm.SVR(kernel="linear") |
Modelos lineales parecidos
| | Sklearn (CPU) | RAPIDS (GPU) | |————————————|——————————————-|———————————| | Support Vector Machines Classifier | | cuml.svm.SVC(kernel=”rbf”) | | Support Vector Machines Regressor | | cuml.svm.SVR(kernel=”rbf”) | | Nearest Neighbors Classification | | cuml.neighbors.KNeighborsClassifier | | Nearest Neighbors Regression | | cuml.neighbors.KNeighborsRegressor |
Clasificación o Regresión?
4. Descenso por Gradiente (SGD)
