Channel Avatar

Data School @UCnVzApLJE2ljPZSeQylSEyg@youtube.com

None subscribers - no pronouns set

Are you trying to learn data science so that you can get you


04:48
Course outline: "Master Machine Learning with scikit-learn"
01:54
Course overview: "Master Machine Learning with scikit-learn"
01:40
Introduction to model ensembling
02:20
How to save a scikit-learn Pipeline with custom transformers
03:52
Should I shuffle samples with cross-validation?
04:16
Cost-sensitive learning in scikit-learn
01:50
scikit-learn vs Deep Learning
07:31
How to read the scikit-learn documentation
02:47:31
My top 50 scikit-learn tips
24:40
21 more pandas tricks
07:49
Adapt this pattern to solve many Machine Learning problems
05:07
Tune multiple models simultaneously with GridSearchCV
03:38
Access part of a Pipeline using slicing
04:07
Tune the parameters of a VotingClassifer or VotingRegressor
04:32
Ensemble multiple models using VotingClassifer or VotingRegressor
04:08
Create feature interactions using PolynomialFeatures
02:16
Speed up GridSearchCV using parallel processing
06:59
Use OrdinalEncoder instead of OneHotEncoder with tree-based models
03:11
Passthrough some columns and drop others in a ColumnTransformer
04:07
Drop the first category from binary features (only) with OneHotEncoder
01:36
Estimators only print parameters that have been changed
02:11
Load a toy dataset into a DataFrame
02:18
Get the feature names output by a ColumnTransformer
02:08
Create an interactive diagram of a Pipeline in Jupyter
01:56
Most parameters should be passed as keyword arguments
01:23
Don't use .values when passing a pandas object to scikit-learn
02:29
Add feature selection to a Pipeline
04:09
Use FunctionTransformer to convert functions into transformers
03:40
Use AUC to evaluate multiclass problems
05:12
Shuffle your dataset when using cross_val_score
02:02
Four ways to examine the steps of a Pipeline
01:57
Vectorize two text columns in a ColumnTransformer
01:45
Save a model or Pipeline using joblib
02:38
Two ways to impute missing values for a categorical feature
04:26
Use stratified sampling with train_test_split
01:35
Prune a decision tree to avoid overfitting
03:54
Visualize a decision tree two different ways
01:21
Display the intercept and coefficients for a linear model
01:46
Use the correct methods for each type of Pipeline
01:52
Compare multiple ROC curves in a single plot
03:04
Plot a confusion matrix
04:40
Important tuning parameters for LogisticRegression
02:36
Display GridSearchCV or RandomizedSearchCV results in a DataFrame
04:42
Try RandomizedSearchCV if GridSearchCV is taking too long
07:02
Use cross_val_score and GridSearchCV on a Pipeline
04:37
Three reasons not to use drop='first' with OneHotEncoder
02:52
HistGradientBoostingClassifier natively supports missing values
03:03
Examine the intermediate steps in a Pipeline
02:55
What is the difference between Pipeline and make_pipeline?
05:50
Impute missing values using KNNImputer or IterativeImputer
02:36
Set a "random_state" to make your code reproducible
03:02
Add a missing indicator to encode "missingness" as a feature
03:12
Use Pipeline to chain together multiple steps
03:17
Handle unknown categories with OneHotEncoder by encoding them as zeros
05:16
Encode categorical features using OneHotEncoder or OrdinalEncoder
04:04
Four reasons to use scikit-learn (not pandas) for ML preprocessing
04:20
Use "fit_transform" on training data, but "transform" (only) on testing/new data
02:44
What is the difference between "fit" and "transform"?
04:07
Seven ways to select columns using ColumnTransformer
03:24
Use ColumnTransformer to apply different preprocessing to different columns