Lecture: “Visualisation Techniques in Data Mining & Applications”

Lecture: “Visualisation Techniques in Data Mining & Applications”

A guest lecture on Visualisation Techniques in Data Mining for real-world applications is given by Gerasimos Antzoulatos, i2s Data Analyst, at the Department of Information and Communication Systems, at University Mittuniversitetet – Mid Sweden University (http://www.miun.se/).

The main idea of the lecture was related to how the data analysts and also end-users in Aquaculture sector could visually investigate data, so as to gain insights and useful knowledge for further analysis. The lecture focused on real-world applications and the findings of AQUASMART European project (http://www.aquasmartdata.eu/) are exhibited. Generally, visualization techniques classified into two broad classes, namely, Data visualization tools and techniques and Visual data mining tools and techniques. Techniques of the former category help Data Analysts as well as aquafarmers to create two and three-dimensional pictures of business data that can be easily interpreted to gain knowledge and insights into those data sets. With data visualization, farm managers act as the data mining or pattern recognition engine. By visually inspecting and interacting with the two- or three-dimensional visualization, they can identify the interesting (nontrivial, implicit, perhaps previously unknown and potentially useful) information or patterns in the business production data set. Aquasmart project provides to aquafarmers the opportunity to investigate their data by utilizing the Free Data Exploration Tool (http://www.aquasmartdata.eu/data-exploration-tool/).

 

Screenshot_1Screenshot_2 Screenshot_1

Decision making and new business opportunities prediction

   On the other hand, techniques of the latter category, help mainly Data Analyst to create visualizations of data mining models hence, gain knowledge and insights into the patterns discovered by the data mining algorithms that cater on the decision making and new business opportunities prediction. With visual data mining tools, Data Analyst can inspect and interact with the two- or three-dimensional visualization of the predictive or descriptive data mining model to understand (and validate) the interesting information and patterns discovered by the data mining algorithm in the aquaculture sector. Furthermore, in the context of the lecture, the significance of data visualization techniques of sharing and interpreting in a meaningful way the results of data mining analysis to the end-users of the aquaculture sector, is highlighted. Finally, there is a presentation of the objectives and the modules of Aquasmart program focusing on Data Analytics part.