Torus hosted a Hackathon

Here at Torus-Actions we take great pleasure in sharing our passion for artificial intelligence and wish to enlighten others on this topic, whether they are just getting started or already have experience in this field. Therefore, we were thrilled to achieve this goal by hosting a Hackathon, a gathering of programmers and software developers for the purpose of collaboration.

The participants, mainly students from different universities and institutions, had the chance to discover artificial intelligence as well as its place in the professional workplace during the two-day event.

The first day was focused on data, a vital part of artificial intelligence. Raw data is often unusable, hence the participants were faced with cleaning and organizing Real Estate data in order to try to predict the value of a property based on specific characteristics.

On the second day, the work revolved around models. Two projects were brought to light, one to predict bitcoin values and the other to recognize written numbers. For the first one, participants had to create a model that would consider the importance of their data’s temporal aspect. In the case of the second project, participants discovered a well-known machine learning model, K-Nearest Neighbours (KNN). However, new problems were unveiled due to the fact that the model analyzes images, and some parts of the image were built into the algorithm to improve performance.