Free data and AI check

Check your data and AI foundations.

Review data quality, visualization, model training, evaluation, bias, privacy, and responsible use of generative AI. You will receive an indicative AI1-AI4 result.

7 min: Typical time
12: Questions
AI1-AI4: Level range

Question 1 of 12 0 answered

1 Data basics

What is a dataset?

A An organized collection of related observations or records B One guaranteed conclusion C A computer monitor D A password manager

2 Statistics

What is the mean of 2, 4, and 9?

A 3 B 5 C 7 D 15

3 Visualization

Which chart is usually suitable for comparing values across a few categories?

A Bar chart B Unlabeled decorative image C Password field D Audio waveform only

4 Machine learning

In supervised learning, what does the training data commonly include?

A Examples paired with target labels or values B Only unlabeled random noise C No observations D A finished legal contract

5 Evaluation

Why separate training and test data?

A To estimate performance on examples not used to fit the model B To make the dataset look larger C To guarantee perfect predictions D To hide every error

6 Data quality

What is missing-data analysis concerned with?

A Whether absent values could distort conclusions or model behavior B Only the color of a spreadsheet C How quickly a monitor turns on D Whether every row has the same font

7 Reasoning

Why does correlation not by itself prove causation?

A An association may reflect confounding, reverse direction, or coincidence B Correlations contain no numbers C Causation is never studied D All variables are identical

8 Overfitting

What is overfitting?

A Learning training details so closely that performance suffers on new data B Using too little electricity C Compressing a file D Drawing a trend line by hand

9 Classification

What does a false positive mean?

A The model predicts the positive class when the true class is negative B The model predicts every case correctly C The data has no labels D The computer is switched off

10 Bias

Where can unfair model outcomes originate?

A Data, labels, objectives, features, design, or deployment context B Only from damaged hardware C Nowhere if an equation is used D Only from user interface colors

11 Generative AI

What is the safest way to use a generated factual answer for an important decision?

A Verify key claims against reliable sources and relevant expertise B Assume fluent wording guarantees accuracy C Remove all uncertainty labels D Share it without reading

12 Privacy

What does data minimization mean?

A Collecting and retaining only data needed for a defined purpose B Publishing every available field C Keeping all data forever D Removing access controls

Before you begin

Use your own knowledge.

Choose one answer for every question.
Do not use calculators, compilers, search, or reference tools.
There is no timer, so read naturally.
You can retake the test as your skills improve.

This is an indicative self-assessment, not an accredited certificate or formal examination. It samples selected foundations and cannot measure every part of the subject.