Scribd
Upload
18 views5 pages

Understanding Data Types and Attributes

This document provides an overview of key concepts for understanding data, including: - The two broad types of variables are qualitative and quantitative. Nominal and ordinal variables are qualitative, differing in whether categories can be ranked. - Attributes on an interval scale have a natural zero and meaningful differences. - Dimensionality and sparsity are important characteristics to consider in a data set. - Structured, unstructured, and semi-structured are three types of data structures.

Uploaded by

ntozakhelunathi7
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
18 views5 pages

Understanding Data Types and Attributes

This document provides an overview of key concepts for understanding data, including: - The two broad types of variables are qualitative and quantitative. Nominal and ordinal variables are qualitative, differing in whether categories can be ranked. - Attributes on an interval scale have a natural zero and meaningful differences. - Dimensionality and sparsity are important characteristics to consider in a data set. - Structured, unstructured, and semi-structured are three types of data structures.

Uploaded by

ntozakhelunathi7
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

Tutorial 2: Getting to know your data

1. Define what “data” refers to. A collection of data objects and their attributes.

2. Explain what an “attribute” represents. A variable property or characteristic of an object.

3. Name the two broad types of variables that may be present within any given set of data.
Qualitative, quantitative

4. Discuss the similarities and differences between nominal and ordinal variables. Both relate to
qualitative attributes with varying states or categories. However, for nominal variables, there is
no natural order to the categories and they cannot be ranked in any way, whereas the opposite
is true for ordinal variables.

5. Which of the following statements are correct regarding an attribute that is on the interval
scale? You may choose more than one.
a. There is no meaningful zero, and ratios don’t make sense
b. There is a natural zero, but differences are meaningless.
c. There is no meaningful zero, but differences are meaningful.
d. There is a natural zero, and ratios are meaningful

6. Discuss how discrete and continuous attributes differ? Discrete attributes have only a finite or
countable set of values whereas continuous attributes can take on an infinite number of
possibilities in the real number system, usually in some interval.

7. When would you reduce dimensions in your data? When you have a large set of features with
similar characteristics

8. A binary variable with states that are not equally important may be described as…
a. Symmetric
b. Asymmetric
c. None of the above

9. Briefly discuss the characteristics that need to be considered for any given data set.
Dimensionality: having to deal with a large number of features can become problematic,
particularly if the number of attributes exceeds the number of observations.
Sparsity: sometimes most of the data features have high percentage of zeros.
Resolution: It is frequently possible to obtain data at different levels of resolution, and often the
properties of the data are different at different resolutions (or any related answer).

10. Which of the following refers to a type of record data?


a. Transaction or Market Basket Data
b. Graph-Based Data
c. Both A and B correct

11. Briefly describe what time series data refers to. Time series data is a special type of ordered,
sequential data in which each record has a time stamp and the time interval between records is
regular i.e. a series of measurements recorded over time at regular intervals

12. Briefly discuss the 3 different types of data structures?


Structured data is considered as the most traditional form of data – it can be presented in a
tabular format and is straightforward to analyse.
Unstructured data is information that either does not have a predefined data model or is not
organised in a pre-defined manner.
Semi structured data is a form of structured data that does not conform with the formal
structure of data models associated with relational databases or other forms of data tables, but
nonetheless contain tags or other markers to separate semantic elements and enforce
hierarchies of records and fields within the data. Therefore, it is also known as self-describing
structure.

13. Which of the following might be considered correct for the ordered process of Data Mining?
a. Infrastructure, Exploration, Analysis, Interpretation, Exploitation
b. Exploration, Infrastructure, Analysis, Interpretation, Exploitation
c. Exploration, Infrastructure, Interpretation, Analysis, Exploitation
d. Exploration, Infrastructure, Analysis, Exploitation, Interpretation

14. Which of the following may be described as a process in which intelligent methods are applied
to extract hidden patterns and relationships?
a. Warehousing
b. Data Mining
c. Text Mining
d. Data Selection

15. What does the acronym KDD stand for?


a. Knowledge Discovery in Databases
b. Knowledge Discovery Dimension
c. Knowledge Data Definition
d. Knowledge Data Dimension

16. What are the functions of Data Mining?


a. Association and correctional analysis classification
b. Prediction and characterization
c. Cluster analysis and Evolution analysis
d. All of the above

17. Which one of the following statements about the mean is incorrect?
a. It is sensitive to extreme values (outliers)
b. It is a single value that is useful for describing a data set
c. It is always the best way of measuring the centre of the data
d. None of the above

18. Which one of the following statements about the median is incorrect?
a. It is sensitive to extreme values (outliers)
b. The median is often referred to as “the middle”
c. It is robust
d. None of the above

19. Which one of the following statements about the mode is incorrect?
a. It is the most common value
b. It is not affected by outliers
c. There may not actually be a mode
d. None of the above

20. A data set which has two or more modes is called ……….
a. Unimodal
b. Symmetric
c. Multimodal
d. None of the above

21. Which one of the following statements about perfectly symmetric, bell-shaped data is incorrect?
a. The highest point on the curve represents the mean
b. Its standard deviation depicts the bell curve's relative width around the mean
c. Median=mean=mode
d. None of the above
22. Which one of the following statements about positively skewed data is correct?
a. There are outliers/extreme values on the upper end of the scale
b. The tail of the distribution lies on the left side
c. The mean is the best measure of centrality for such data
d. None of the above

23. Which one of the following statements about negatively skewed data is incorrect?
a. There are outliers/extreme values on the lower end of the scale.
b. The value of the mean is the greatest one followed by median and then by mode
c. The median is the best measure of centrality for such data
d. None of the above

24. Which of the following measures are appropriate for assessing the dispersion of an asymmetric
numeric variable?
a. Range
b. Percentile
c. Variance
d. All of the above

25. For data containing multiple continuous variables, the spread or variability of the data can be
represented using…
a. Range
b. Covariance matrix
c. Interquartile range
d. None of the above

26. Which one of the following statements regarding the measured covariance between two
variables is correct?
a. A value near zero denotes that the two variables do not have a linear relationship
b. It measures the direction of a relationship between two variables
c. Covariance is different from the correlation coefficient
d. All of the above

27. Which one of the following statements about the correlation coefficient is incorrect?
a. Correlation coefficient indicates how strongly two variables are (linearly) related
b. A positive correlation means that the two variables move together in the same direction
while a negative correlation means they move inversely.
c. A perfect positive correlation means that the correlation coefficient is exactly 1
d. None of the above

28. State what type of data visualization would be best to use in order to:
a. Explore the relationship between 3 different quantitative variables 3-dimensional scatter
plot / bubble plot
b. Compare the shape of the distributions of 4 different quantitative variables multiple box
plots
c. Identify outliers box plot
d. Partition spatial data in to regions of similar values contour plot
e. Explore the relationship between 2 different quantitative variables plus a 3rd categorical
variable scatter plot

Common questions

Powered by AI

Interval scales have no meaningful zero point, meaning ratios are not meaningful; differences can be measured (e.g., temperature in Celsius). Ratio scales have a natural zero point where both differences and ratios are meaningful, such as height or weight (e.g., 0 kg means no weight).

Covariance indicates the direction of a linear relationship between two variables, but its value is not standardized, making it difficult to compare across datasets. The correlation coefficient, however, is standardized to a range from -1 to 1, allowing comparability, and quantifies the strength and direction of the linear relationship .

Structured data is highly organized and easily searchable in tables or columns, such as a database or spreadsheet. Unstructured data lacks a predefined structure, making it more complex to analyze; examples include text documents, images, and videos. Semi-structured data does not conform to a strict data model but uses tags or markers to define elements and hierarchies, such as XML or JSON files .

Nominal variables are qualitative and have categories without a natural order, such as colors (red, blue, green). Ordinal variables also have categories but with a meaningful order, such as customer satisfaction ratings (poor, fair, good, excellent).

Different visualizations are utilized based on the data characteristics; for example, a 3-dimensional scatter plot is used to explore relationships between three quantitative variables, multiple box plots for comparing distributions, box plots for identifying outliers, contour plots for partitioning spatial data, and scatter plots with a categorical variable to explore two quantitative variables .

Data sparsity affects storage and processing efficiency, as a high percentage of zero values can complicate analyses. High dimensionality can lead to the 'curse of dimensionality,' where the number of features complicates model fitting and interpretation. Varying resolution levels affect the analysis detail, potentially masking or revealing different patterns in the data .

Time series data, defined by ordered sequences with time stamps and regular intervals, can be used to model and predict future events by identifying trends, seasonal patterns, and cyclical fluctuations. Applications include stock market predictions, weather forecasting, and economic trend analysis .

Dimension reduction is crucial for simplifying data analysis by reducing the number of variables, thereby decreasing computational cost and avoiding multicollinearity. It is particularly recommended when there's a large feature set with similar characteristics, which may overwhelm traditional analysis techniques .

The median is preferred over the mean in skewed data as it is less sensitive to extreme values (outliers). In positively skewed data, outliers are in the upper range, causing the mean to be higher than the median, while in negatively skewed data, outliers are in the lower range, causing the mean to be lower than the median .

KDD (Knowledge Discovery in Databases) is a broader process that encompasses the entire discovery of useful knowledge, whereas data mining is a specific step within KDD focused on extracting patterns from data. Core functions of data mining include association and correlation analysis, classification, prediction, cluster analysis, and evolution analysis .

You might also like