To understand patterns and structure in a dataset before modeling

To automatically deploy a trained model to users

To encode text labels into database primary keys

Handling missing values, duplicates, and inconsistent entries

Choosing which chart colors to use for a report

Selecting the final hyperparameters for a model

Quantitative data is numeric, qualitative data describes categories or qualities

Quantitative data is always collected by surveys

Qualitative data can only be visualized with scatter plots

Two variables can move together without one actually causing the other

Correlated variables must always be plotted as scatter plots

A correlation coefficient can never be calculated for real data

They can distort statistics like the mean and skew model training

They always increase the size of the training set

They automatically improve model accuracy

The sum of all values divided by the count

It makes patterns and trends easier to spot quickly

It removes the need to clean the data first

Using a misleading axis scale that exaggerates differences

Choosing colors that are easy to distinguish

Poor quality input data leads to unreliable model results

Models always produce perfect results regardless of data quality

Only the visualization step affects the accuracy of a model

An input variable used by the model to make predictions

The final accuracy score of a trained model

A row of the dataset containing the target label only

The outcome or value the model is trying to predict

A hyperparameter chosen before training begins

Supervised learning uses labeled data, unsupervised learning does not

Supervised learning never requires any data cleaning

Unsupervised learning always predicts a numeric target

Grouping customers into segments based on purchase behavior

Predicting house prices from labeled sale records

Classifying emails as spam using labeled examples

The model learns the training data too closely and fails to generalize to new data

The model is too simple to capture patterns in the data

The model was trained on too little data to run at all

Simplify the model or use more training data

Remove the test set entirely before evaluation

Increase the model's complexity without limit

To document what each column or field in a dataset represents

To calculate the correlation between two variables

Filling in missing values using estimates such as the mean or median

Deleting the entire dataset when any value is missing

Converting numeric data into categorical bins

Data Science Quiz