Machine learning activity detection using ML.Net
Anca Alexan, Alexandru Iulian Alexan, Stefan Oniga · 2020
Our living environment is becoming more and more aware of our presence and starts to react and interact with us. The smart house has long moved from concept to reality., as our homes are now digitalized with all sorts of smart devices. Since we now have more data available the ever., an important part is to be able to analyze the data and provide the user with useful information. Neural networks are a good way of providing systems capable of handling large volumes of data that are able to learn based on users' data. Our proposed neural network system consists of a .NET software application that processes data from the CASAS activity detection system. The software application uses the ML.NET NET machine learning framework for predicting the users' activity. The ML.NET framework was chosen since its open-source and cross-platform.