7TRAN033W ReferredDeferred Statistics Coursework

7TRAN033W
REFERREDDEFERRED
STATISTICS
COURSEWORK

Question 1(a): Decision Tree
A decision tree tool commonly referred to as a decision support system modelled as a tree
like structure. It is widely used in businesses for evaluating different options and thus is helpful
in choosing the best possible or effective courses of action. The following is the decision tree
depicting alternative options available to the government while attempting to reduce the number
of trucks that are producing excessive carbon emissions. Each alternative is presented through
their consequent outcomes.
Figure 1: Decision Tree
(Created by learner)
Through the above chart, it has been identified that there are three options available to the
government with regards to reducing the number of trucks with excessive carbon emissions that
is, do nothing, increase fine and run advertising campaign. While going with doing nothing, the
situation would remain the same with 19% carbon emission. When going with the option of
increasing fine, there are 30% chances that the carbon emission would get reduced to 10%
otherwise 70% chances are of increasing carbon emissions to 25%. The third option available
with the government involves running advertising campaign through television or trucker
magazines. If television advertising would be opted for, then the carbon emission would get
reduced to 17% whereas going with the option of running advertising campaign through trucker

there would 40% and 60% probability of reduction in carbon emission to 12% and 18%
respectively.
Question 1(b): The Decision Tree Analysis
Option 1 – Do nothing = 19%
Option 2 – Increase fine = 0.3 * 10% + 0.7 * 25% = 3 + 17.5 = 20.5%
Option 3 – Run Advertising Campaign
- Television = 17%
- Trucker Magazines = 0.4 * 12% + 0.6 * 18% = 4.8 + 10.8 = 15.6%
Different attitudes towards the risk and selection of option
Risk Taking = A risk taking government will go with increasing fine where the probability of
getting maximum reduction in carbon emission to 10% is just 30% while the probability of
increase in carbon emission to 25% is 70%. Therefore, going for this option with the aim to
reduce carbon emissions to 10% is highly risky element because of the lower probability of
occurring event.
Risk Averse = A risk averse government who can’t take risk of any kind will go for running
advertisement through television with the 100% probability of reducing carbon emission by 2%.
Risk neutral = A risk taking government will go with running the advertising campaign through
Trucker magazine which would result in maximum reduction in carbon emissions.
Question 1(c): The Assumptions in relation to the analysis
There are no budget constraints pertaining to the selection of option for reducing carbon
emission.
- There is an immediate requirement for taking action by the government pertaining to
reduce carbon emission.
- The individual options are independent of each other.
- The entire training set is taking into account as root.
- Distribution of records is done recursively based on an attribute values.