Skip to content
Smartable Academy
Smartable tools

← Three Kinds of Guess, and Their Data

Your progress

Checking your sign-in…

Who's learning?

Each profile keeps its own read lessons, badges, streak, and practice scores on this device.

Lesson 8 of 9

Track quiz

Show on a big screen

8 min

Text size

Before the questions, here is each idea in one line to keep.

One line to keep from each idea of this track
IdeaThe line to keep
Guess a numberRegression learns from past numbers and guesses a new number.
Group with no labelsClustering finds groups in things with no labels. What it compares decides the groups.
Three kinds of guessAsk what comes out. A named group is classification. A number to guess, like a price, is regression. Groups found with no labels are clustering.
What a dataset holdsA dataset is a set of records. Each record is described by attributes. One way to name a dataset is by what it holds.
Structured or unstructured dataStructured data fits rows and columns. Unstructured data does not. A dataset with both is hybrid.
Train, check, testTrain on one part. Improve with a second. Test on a third it never saw.

A program learns from past petrol bills. It guesses next month's bill in rupees. What is that?

  • Regression, because the answer is a number
  • Classification, because it sorts the bills
  • Clustering, because bills are grouped by month

A line drawn through the cycle dots misses some of them. What does that tell you?

  • The line is wrong, because it should touch every dot
  • The cycles were priced wrongly, because dots belong on a line
  • Its guess can be a little off, because real cycles vary

A program groups 200 unmarked leaf photos by how alike they look. What did nobody give it?

  • Photos, because it found its own
  • Labels, because the photos were unmarked
  • A number to guess, because it made groups

Buttons are grouped by size. The same buttons are then grouped by holes. The groups change. Why?

  • The program made a mistake, because groups must stay the same
  • It compared something different, so the groups are different
  • The labels changed, because holes are labels

A program learns from photos marked ripe or unripe. It marks a new photo. Which kind of guess is that?

  • Classification, because the answer is a named group
  • Regression, because ripeness is measured
  • Clustering, because it puts photos together

A program says a match will be won or lost. A learner calls that regression. What is the fix?

  • Nothing, because won and lost are numbers
  • It is clustering, because there are two groups
  • It is classification, because won and lost are named groups

Each row of a rain table is one day. What is one row called?

  • An attribute
  • A dataset
  • A record

A lake's water level is measured every morning for a year. Which kind of dataset is it?

  • Text
  • Time-series
  • Spatial

A folder holds 300 phone videos. Which kind of data is it?

  • Unstructured, because it has no fixed rows and columns
  • Structured, because each video is a file
  • Hybrid, because videos have sound and pictures

A form has a drop-down list and a comment box. What is its data?

  • Structured only, because it is a form
  • Hybrid, because it has both kinds
  • Unstructured only, because the comment is free

A dataset is split in three. Which part gives the last fair score?

  • The training set
  • The validation set
  • The test set

A team tests its model on the photos it trained on. It scores full marks. What can you say?

  • Nothing about new photos, because it saw these answers
  • The model will be right on new photos too
  • The model is wrong, because full marks is too high

You can now name what a model guesses. You can name what it learned from. Next comes Data Science, Vision and Language.

Reveal every answer to finish this lesson.