Neural Networks, C# and telecoms fraud detection final year project

One of the things I have regretted since I left university is that I have not followed up on my Final year Project [I will try and upload it soon, it was pretty smart.]. Bascially I made a system that created Telephone CDR records based off user determined call profiles. The data from these CDR was then passed into a neural network [Created using MATLAB] to see if the call history for a particalular telephone customer was fraudulent or not. It worked really well, and MATLAB was amazing for creating the nerual networks for training and validating.

I promised myself that I would create a little neural network framework in what ever language I was working with at the time so that I could learn more about the algorithms and functions of the network. MATLAB was great but it was easy to hide the detialed understanding of the inner workings of a neural network.

So, soon hopefully I will create a little C# neural network package. Just mainly as a learning exercise. I know there are a lot of resources out there that already do it in c#, but I don't really want to use them for many reasons. Some of them seem convulted, some of them seem very specialised and all of them will not really help me understand neural networks the way that I should, it will be MATLAB all over again.

On the neural network side of things I used a Multi-layer perceptron. Which had an astonishing high true-positive/negative rate.

Basically my next series if posts will be about my Final year project and the work that I did with neural networks. Additionally, I will try and talk about some of the c# aspects of me learning about neural networks again.

About Me: Paul Kinlan

I lead the Chrome Developer Relations team at Google.

We want people to have the best experience possible on the web without having to install a native app or produce content in a walled garden.

Our team tries to make it easier for developers to build on the web by supporting every Chrome release, creating great content to support developers on, contributing to MDN, helping to improve browser compatibility, and some of the best developer tools like Lighthouse, Workbox, Squoosh to name just a few.