Split tests (e.g., hold-out) are used for the hyperparamet…
Split tests (e.g., hold-out) are used for the hyperparameter-tunning and predictive performance evaluation. Additionally, split test design depends on the data (e.g., cross-validation for Independent and identically distributed data). The simplest split test design is the hold-out method During the training step in the hold-out split-test for a supervised learning (SL) model, Step 1. The SL model trained with (1)__________(a. total train, b. sub-train, c. valid, d. test; 1 point) dataset. Step 2. Trained model predicts labels in (2)__________(a. total train, b. sub-train, c. valid, d. test; 1 point) dataset. Step 3. Repeat steps 1 and 2 with different hyperparameters (e.g., h-parameter 1={1,2} h-parameter 2= {1,2} = 4 times) Step 4. Finding out the best hyperparameters for the SL model During the test (i.e., inference) step in the hold-out split-test, Step 5. Training SL model with the optimal hyperparameters and (3)__________(a. total train, b. sub-train, c. valid, d. test; 1 point) dataset. Step 6. Predict labels in (4)__________(a. total train, b. sub-train, c. valid, d. test; 1 point) set and evaluate it.
Read Details