Q1:

1. Select a single metric from the data (e.g. age)

2. Design a MATLAB script to compute the population standard deviation and the Interquartile Range of this metric. Annotate each line of your script to describe what it does.

3. Imagine that this data is a sample of a wider population. Create a script to compute the sample standard deviation. Annotate each line of your script to describe what it does.

4. Choose any data column from the responses (with number values), and use MATLAB to produce a boxplot of this data, you must ensure you label this data appropriately. Annotate your boxplot using MATLAB to indicate what each element means. Annotate each line of your script to describe what it does.

Make sure you clearly show how you got to your results, include: -

- a screenshot of each MATLAB script that you created (including annotations)

- an image of the resulting graph including annotations (made using MATLAB).

Q2:

This assessment concerns the evaluation of a machine learning system.

The local Council is considering purchasing a waste sorting system, for the purposes of recycling. This takes a photograph of each object in turn, before applying a computer vision algorithm to sort each object into the relevant category. The machine is capable of sorting into the following categories: plastic bottles (B), tin cans (T), glass jars (J) and everything else (X).

To measure the performance of the waste sorting system, the council tested the system with a sample of 30 items, which were in turn sorted by the machine and given a Label.

By hand and showing your working:

(a) Compute the Confusion Matrix for the Machine, based upon the sample above

(b) Using the Confusion Matrix, or otherwise, compute the Accuracy of the waste sorting system, based on the sample above.

(c) The Council is considering using the Accuracy figure based on the sample as a means for deciding if it should purchase the waste sorting system. Explain why this might not be appropriate.

(d) The Council considers that they will use the system purely as a machine for identifying and extracting plastic bottles (and which will reject everything else). On that basis, compute the Precision, Recall and F1-Score using the Confusion Matrix, or otherwise.

Data set