logistic regression

abalone dataset · predicting sex (M/F) from physical measurements

regularization

λ 0.01

learning rate 0.01

epochs 5000

train — gradient descent with the λ from the slider. search & train — finds the best λ automatically via 5-fold cross-validation, then retrains with it. this is the correct workflow.

λ min (10^x) -6

λ max (10^x) 1

grid steps 30

cv epochs 5000

loss evolution

train vs validation binary cross-entropy — lower is better

accuracy evolution

train vs validation accuracy — fraction of correctly classified samples

predicted probabilities

each dot is one abalone — x is P(male), rows are actual sex

feature weights

weight magnitude after training — positive = more likely male

weight stability across folds

distribution of each feature's weight across 5 CV folds — wide spread reveals instability

grid search — λ vs validation accuracy

5-fold cross-validation across λ values · best λ used to retrain