CFA Level I · CFA Level I Exam · Introduction to Financial Data Science
To assess how well a model will generalize, an analyst splits the data into a training set, a validation set, and a test set. The test set is most appropriately used to:
The test set is used to give a final unbiased evaluation of the chosen model. Parameters are fitted on the training set and hyperparameters are tuned on the validation set, so the test set remains untouched and gives an honest estimate of out-of-sample performance.
- Atune the model's hyperparameters
- Bprovide a final unbiased evaluation of the chosen modelCorrect
- Cfit the model's parameters to the data
Explanation
The training set fits parameters, the validation set tunes hyperparameters and compares models, and the test set gives a final out-of-sample evaluation. Using the test set for tuning would compromise its independence and bias the estimate of performance.
Did you get it right without looking?
One question tells you little. A timed set on Introduction to Financial Data Science shows your real accuracy, how long you take and where you lose marks.
More Introduction to Financial Data Science questions
- A model achieves a very low error on its training data but a much higher error on a validation data set. This result most likely indicates t…
- A fund uses a model that classifies firms as likely to default or not, using labeled historical data on past defaults and firm characteristi…
- A model achieves very low error on its training dataset but produces much larger errors on a validation dataset. The model is most likely su…
- In a machine learning project, the data are divided into a training set, a validation set, and a test set. The test set is most likely used …
- An analyst computes term frequency-inverse document frequency (TF-IDF) for the word "covenant" in a loan filing. The word appears 6 times in…
- An analyst's dataset of customer incomes contains several extreme values far above the rest, and the analyst wants to reduce the influence o…