[FIXED] How Should I Write a Neural Network for Prediction on this CSV Data?

Issue

I am trying to do a university project, and my professor asked me to do a Neural Network to predict servers' behavior in this CSV data. https://drive.google.com/file/d/1Su8CnliKcclYxmSOlEg7zu_73X2WO9J4/view?usp=sharing

nodeid = id of a server
node_cpu_usage = server cpu usage
node_memory_usage = server memory usage

And the task is to train a model on this server's CPU and memory usage to predict how they will behave in the future. What is your suggestion?

Solution

This falls under time anomaly detection using 2 input variables: (CPU, & memory). In NN world, the time series recurrence is trained using LSTM. If you Google with search terms time anomaly detection using LSTM or multivariate anomaly detection, you will get lots of academic work about this type of implementations, and some sample reference implementations. Not all these implementations use neural network; but should give you an idea about exploring and pre-processing your data.

Some references for help:

• https://becominghuman.ai/time-series-and-how-to-detect-anomalies-in-them-part-ii-bde9e69d0aaf
• https://datascience.stackexchange.com/a/77872/38183

Answered By - Azhar Khan

Answer Checked By - Marilyn (PHPFixing Volunteer)