Terminology

Question 1

Overfitting

Answer 1

A phenomenon where a model does very well on training data but does poorly during validation or on new data

Question 2

Underfitting

Answer 2

When a model does very poorly because it failed to capture important features or distinctions even on the training set

Question 3

Validation set

Answer 3

A subset of the training data set used to test the early accuracy of a model during the tuning stages.

Question 4

Training set

Answer 4

A data set used to fit the model.

Question 5

Testing set

Answer 5

data set used to provide an unbiased evaluation of a model.

Question 6

Model fitting

Answer 6

Approximation of data to a target function.

Question 7

Decision tree regression

Answer 7

Similar to decision tree but used to find a continuous value and mean squared error is used to determine the number of splits.

Question 8

Imputation

Answer 8

The filling of missing values in a dataset

Question 9

Categorical attributes

Answer 9

Values that fall into a set of ‘categories’

Question 10

What are techniques are used to deal with categorical data?

Answer 10

1 ) dropping categorical columns.

2) Label encoding
3) One-hot encoding

Question 11

Label encoding

Answer 11

Assigning a unique integer to a categorical value

Question 12

One-hot encoding

Answer 12

creation of new columns for each unique categorical

Question 13

Pipeline

Answer 13

automation of workflow to bundle preprocessing and modeling together

Question 14

Cross validation

Answer 14

Subsets of training data used to provide a more accurate reading

Question 15

Variance

Answer 15

How different the results are when the model is tested on new data sets.

Question 16

Data leakage

Answer 16

When training data contains some feature related to the target in such a way that the model has great accuracy on the training data but does poorly to unseen data