User interface language: English | Español

Date November 2018 Marks available 4 Reference code 18N.2.HL.TZ0.8
Level HL Paper 2 Time zone no time zone
Command term Explain Question number 8 Adapted from N/A

Question

A parasite is destroying olive trees in a region in southern Europe. This area is monitored by taking aerial digital images at midday each day. A neural network is then used in the analysis of these images.

Genetic algorithms follow an iterative process in developing a solution to a problem.

Outline the steps involved in this iterative process.

[6]
a.

Outline why neural networks need to be trained before they can be used in this analysis.

[4]
b.i.

Explain why supervised learning may be preferred to unsupervised learning for training these networks to make predictions on the progress of this parasitic disease.

[6]
b.ii.

Many companies now use an automated response system when answering basic telephone queries from customers. This system uses speech analysis together with recorded messages.

Explain why the content of the messages given by these systems must be carefully chosen.

[4]
c.

Markscheme

The initial population set is chosen randomly/pseudo randomly;
A fitness function is applied to each population;
The fittest members are selected for the next stage;
Genetic operators are applied;
Such as crossover / mutation;
The process is repeated until;
An acceptable level of fitness is found / a plateau is reached / a maximum number of iterations has been reached;

a.

Network networks initially have a randomly chosen set of weights;
Which will not solve any problem;
These weights must be altered;
Until they fulfil their objective;

b.i.

Award up to [6 max].
Award [3 max] for considerations linked to each of the two forms of learning.
With supervized learning the processing of the image is guided (supervised) by the output dataset to the purpose of classifying certain features of the image;
For example, to the value of the pixel (grey or colour) one can assign likelihood of belonging to neighbouring values of greys/colours;
And this can be defined in advance as a way to represent a potential value for an infected tree;

However, with unsupervised learning the processing of the image is not guided in any way / the output is not known;
And the purpose is to discover and build clusters that have common characteristics;
Therefore, this may clearly help in building up the main regions, but not specifically on focussing on the infected trees;

b.ii.

Award up to [4 max].
Questions/requests must be short so repetition is not required / so they are immediately understood;
And chosen so that there is a restricted response set;
That can be recognised by the system / that can be stored by the system;
The requests must lead to a satisfactory end;
Or customers will get frustrated / hang up / the company will lose business etc.;

c.

Examiners report

[N/A]
a.
[N/A]
b.i.
[N/A]
b.ii.
[N/A]
c.

Syllabus sections

Option B: Modelling and simulation » B.4 Communication modelling and simulation
Show 29 related questions
Option B: Modelling and simulation

View options