Skip to 0 minutes and 1 secondIn this presentation, we're going to look at the art of data visualisation. When we talk about data visualisation, we're talking about taking some findings from data analytics and making it compelling in a visual way by attracting the information up to a level where people can easily understand it. Because we are doing it for this purpose, we need to consider the user or the audience for the visualisation. It's very important.
Skip to 0 minutes and 30 secondsIf we take the big data machine learning algorithms that we've seen before, such as regression analysis, we very quickly understand that the best way to show this is to graph it, because even if it's the case of the researcher looking at their own work, that is the level that they want to see it at to understand if they've been accurate or not. Another example from machine learning is around classification. You have a number of different areas where your classing your data into. And in this example, we probably want to use some sort of a different visualisation, such as a bubble graph, where we can see two, three, or even four different variables all shown in the data.
Skip to 1 minute and 14 secondsAlso in terms of clustering, we'd like to look at data again in a graph format. It seems to be the easiest way for the user to understand it. Outside of the concept of machine learning and very specific analytics, we've also got maybe more general analytics that is undertaken by people in other spheres of work. And in this case, they might look to use a number of different visualisations techniques. First of all, for example, you could be looking at understanding whether you'd like to visualise interactively to allow people to look at the data in different ways and pull out different themes from that data, or we could do it statically.
Skip to 1 minute and 57 secondsAnd we can show simply a visualisation on a map or a graph. As we can see in this example, the interactive visualisation will allow our user to dive deeper into a data set and find out more things about it. The producer of that visualisation has decided they want it to be interactive, because it leads to a much more impressive user experience. We also can look at static visualisations as well. And in this example, you can see that the benefit of the data being visualised statically and it doesn't change over time.
Skip to 2 minutes and 36 secondsAnother example to consider would be this overlay of a map of transport data in London, where we can see in real time how the impact of changes in the network are impacted on our map. So we're able to understand very quickly what's actually happening with the data. We also might want to consider our user audience as being, for example, a manager. And a manager might want to see information about their area in a dashboard format. It gives them very quickly high level of insight into what's happening, but they're not concerned about drilling down into the data. So in this example of a dashboard, you can see very quickly how the city is operating.
Skip to 3 minutes and 20 secondsAnother key component when delivering good visualisations is the art of using colour. Colour helps set the tone in a visualisation and can be used to describe different variables. And you can very quickly see in this example how the use of colour gives you a quick understanding of the data that's being shown. Furthermore, colour is really important to make it accessible to all types of user. And we to consider these elements in good visualisations as well.
Data analytics and visualisations
In this discussion, Steve provides his thoughts and gives some examples on exploring and modelling data in the health and care sector.
© University of Strathclyde