Page 277 - Artificial Intellegence_v2.0_Class_11
P. 277
Body weight Sweets consumption (in grams) per week Total
(in Kg) < 50 50-150 >150
< 50 27 4 2 33
50-69 25 34 2 61 Marginal
70-89 8 41 19 68 Total
>=90 4 7 27 38
Total 64 86 50 200
Marginal Total Total number of samples
Through the above table we can see that
• 27/50 = 54% people who eat more than 150 g of sweets per week weigh more than 90 kg.
• Similarly, 27/64=42% people who eat less than 50 g of sweets per week weigh less than 50 kg.
• 4/64 = 6% people who eat less than 50 g of sweets per week weigh 90 kg or more.
Hence, this shows that people who eat more sweets tend to weigh more and people who eat less sweets, in all probability,
weigh less. So, the table is able to establish this relationship between sweets consumption and body weight.
Scatterplots
A scatter plot (also called a scatterplot, scatter graph, or scatter diagram) is a type of graph which uses Cartesian
coordinates to display values of mainly two variables in a dataset. It typically, consists of an X-axis (the horizontal axis), a
Y-axis (the vertical axis), and a series of dots. Each dot on the scatterplot signifies one observation from a data set. The
position of the dot on the scatterplot represents its value on the X axis and Y axis respectively.
Scatter plots usually represent large amount of data. Scatter plots are used to inspect the relationship between two
variables. These help to analyse how much one variable affects the second variable. This relationship between the two
variables is called correlation. Scatterplots are used with continuous data. We are generally not concerned about single
observations, but rather about the structure of the whole dataset.
Example: A scatterplot showing relation between number of hours studied and percentage (of marks). Note that to plot
the points on the graph, you will show each one as an ordered pair (hours, percentage).
Study Time (hours) 4 3.5 5 2 3 6.5 0.5 3.5 4.5 5 1 1.5 3 5.5
Percentage 82 82 90 74 40 97 51 75 86 85 62 75 70 91
120
100
80
Percentage 60
40 Outlier
20
0
0 1 2 3 4 5 6 7
Study Time
Regression 275

