Compare supervised and unsupervised learning with examples, algorithms, data requirements and decision rules. Use the dedicated the two learning setups dataset and notebook to reproduce the lesson rather than reading it passively. This page targets the practical intent behind the two learning setups, not repetition of the phrase itself.
Editorial note: Publication metadata for Supervised vs Unsupervised Learning is intentionally unset until this URL is actually released; the lesson already includes its own examples, CSV and notebook.
Frame the prediction problem before choosing a model
Predicting whether a customer will churn is supervised because historical churn labels exist. Grouping customers into behavior segments without predefined labels is unsupervised.
- Supervised learning uses labeled examples to predict targets, while unsupervised learning searches for structure without target labels.
- Classification predicts categories and regression predicts continuous quantities; clustering and dimensionality reduction are common unsupervised tasks.
- The same dataset can support supervised and unsupervised questions depending on the objective.
- Choose the learning setup from the available labels and decision goal, not from which algorithm sounds more advanced.
Build a baseline and a leakage-safe evaluation
For the two learning setups, predict one visible result before each snippet runs—a row count, shape, statistic, ranking or error pattern—then compare it with the printed checkpoint.
# the two learning setups · example 1
from sklearn.linear_model import LogisticRegression
from sklearn.cluster import KMeans
clf = LogisticRegression().fit(X_train, y_train) # supervised
clusters = KMeans(n_clusters=3, random_state=42, n_init='auto').fit_predict(X) # unsupervisedthe two learning setups checkpoint 1: Run the cell and inspect the printed values before changing anything.
# the two learning setups · example 2
import pandas as pd
from sklearn.model_selection import train_test_split
df = pd.read_csv('practice.csv')
target = df.select_dtypes('number').columns[-1]
X = df.drop(columns=[target]).select_dtypes('number')
y = df[target]
X_train,X_test,y_train,y_test=train_test_split(X,y,test_size=.25,random_state=42)
print(X_train.shape, X_test.shape)the two learning setups checkpoint 2: Training and test shapes are printed so the split is explicit.
# the two learning setups · example 3
from sklearn.dummy import DummyRegressor
from sklearn.metrics import mean_absolute_error
baseline=DummyRegressor(strategy='mean').fit(X_train,y_train)
pred=baseline.predict(X_test)
print('baseline MAE=', round(mean_absolute_error(y_test,pred),3))the two learning setups checkpoint 3: The baseline score becomes the minimum standard a more complex model should beat.
# the two learning setups · example 4
from sklearn.linear_model import LinearRegression
model=LinearRegression().fit(X_train,y_train)
pred=model.predict(X_test)
print('model MAE=', round(mean_absolute_error(y_test,pred),3))the two learning setups checkpoint 4: Compare model MAE with the baseline instead of judging the model in isolation.
# the two learning setups · example 5
errors=(y_test-pred).rename('residual')
print(errors.describe().round(3))
print('largest absolute errors:')
print(errors.abs().sort_values(ascending=False).head())the two learning setups checkpoint 5: Residual statistics and the largest errors reveal cases worth investigating.
Inspect errors instead of chasing one score
The same dataset can support supervised and unsupervised questions depending on the objective.
Choose the learning setup from the available labels and decision goal, not from which algorithm sounds more advanced.
Self-check with answers
For the two learning setups, what result should you be able to explain without the code?
Supervised learning uses labeled examples to predict targets, while unsupervised learning searches for structure without target labels.
Which assumption matters most in this the two learning setups exercise?
The same dataset can support supervised and unsupervised questions depending on the objective.
What failure mode would make this the two learning setups result untrustworthy?
Choose the learning setup from the available labels and decision goal, not from which algorithm sounds more advanced.
Repeat the experiment with a documented change
Try it yourself workspace
Edit this the two learning setups starter, copy it, and run it in the downloadable notebook or your local Jupyter environment.
Questions specific to this lesson
What should I establish before tuning the two learning setups?
For the two learning setups, begin with a leakage-safe split and a simple baseline so later tuning has a credible reference point.
How do I check overfitting in the two learning setups?
For the two learning setups, compare training and validation behavior, inspect held-out errors, and repeat the evaluation with a stable split or cross-validation when appropriate.
Which metric should I report for the two learning setups?
Choose the learning setup from the available labels and decision goal, not from which algorithm sounds more advanced.
How should I explain the two learning setups in a portfolio or interview?
When presenting the two learning setups, show the problem definition, baseline, split strategy, metric, one error-analysis finding and one limitation.
Primary references and further reading
For the two learning setups, use these primary references when exact syntax, defaults or formal definitions matter beyond the examples on this page.
