These Data Science questions use examples you can try in a beginner project. Start with the terms you encounter in your own work.
What is the difference between classification and regression?
Classification predicts a category, such as whether an order will be returned. Regression predicts a numeric value, such as delivery time.
What is a confusion matrix?
It counts actual versus predicted classes. In a binary problem it shows true positives, true negatives, false positives and false negatives.
When can accuracy be misleading?
If 98 of 100 cases are negative, predicting negative every time gives 98 percent accuracy while finding zero positives. Inspect errors by class.
What is the F1 score?
F1 is the harmonic mean of precision and recall. It summarizes both metrics and is low when either is low.
What is cross-validation?
Cross-validation fits and evaluates a model across multiple splits of the training data. Each validation fold serves as held-out data for a corresponding fit.
When should I use a time-based train-test split?
Use it when predictions will be made on future observations. Train on older records and evaluate on later records to reproduce that order.
Why should I fit a scaler on the training set only?
Fitting it on all records lets the test set influence the training process. Fit on training data and use the learned transformation on validation and test records.
What is a classification threshold?
It is the cutoff that turns a model score into a decision. Lowering the cutoff generally flags more cases, which can increase recall and reduce precision.
For more course topics, see the ScholarLogic beginner Q&A guide.
