Whether you are a member of Coding Club, an ICT teacher from near or far, just a curious passerby or a web crawler, you are welcome to download, view and use any of our previous sessions.
Any queries please contact me!
During the apocalypse of 2020, we made efforts to continue Coding Club remotely to keep ourselves busy. Using Google Colab, we were able to actually use a far greater range of modules than at school which has led to some interesting projects.
We investigate SIR models as a way of projecting and visualising how an epidemic would spread. Using the key parameters that allow us to understand a disease, we can model its spread through the population.
Contining from the first session, we use data collected so far on daily cases in the UK combined with Scipy's parameter optimisation capabilities to deduce the parameters from the data allowing us to make some rough approximations of the future.
Upon closer investigation, the inaccuracy and varation in the testing strategy capturing all cases exposes our previous model to not closely reflect the situation. In an effort to refine our model to more accurately predict the future, we use fatality data which is more consistently recorded in order to approximate cases and project further into the future and gain a deeper understanding into the current situation.
This is an introduction in to how we can train a neutral network using TensorFlow to get good results. Over time, we can build more and more complex models to solve some of the most challenging problems facing our world.
Maps have been a fundemental method of navigation for thousands of years. But how do we represent vast swathes of spatial data on a computer? Using a graph of nodes and edges, we can efficiently store and apply pathfinding algorithms to road maps.