Preliminary design ideas
First, we designed the ideas on paper as text so that we could get some idea of the project, such as what methods we will use when installing: such as operating system, software and other project related things.
Purpose and goal status
So, we started from scratch with no prior knowledge or experience of machine learning techniques. Project was started as part of Haaga-Helia University of Applied Sciences course “monialaprojekti (ICT-infrastruktuurit) PRO4TN004-3005”. We came up idea with our team Joni Hakala, Tuomo Kuure and our teacher Tero Karvinen, that we could somehow reduce the amount of log information with that is being sent to the servers with help of machine learning, or at least notice the exceptional situations on the server using the log information.
Platforms used in the project
We decided to use Google’s TensorFlow, because it was the most popular neural network artificial intelligence platform now. Giving instructions with code to TensorFlow works in C ++, Swift, Python… We chose Python as we both know at least some coding with Python. As an operating system we are installing some Linux we have not yet chosen which one yet.