Hi , I am fairly new to machine learning in general. But recently I stumbled on an interesting project .

The project is based on the data from [https://www.kaggle.com/shasun/tool-wear-detection-in-cnc-mill](https://www.kaggle.com/shasun/tool-wear-detection-in-cnc-mill)

This dataset is about CNC milling machine and detecting the tool wear of the machine.

So, the dataset is comprised of 18 experiments of CNC milling process , making an "S" shape on a wax block each time. At the end of each of the 18 experiments, the tool condition of the CNC machine is recorded. So in an essence, the tool condition can be detected depending on the settings and the condition of the CNC machinery of each run.

There is 18 csv files, each file representing an experiment run, from start to the end of milling process to make the "S" block. The files contains sensor data such as velocity, spindle position etc. Then there is 1 csv file that record the condition of the tool after each run (worn/unworn).

I would like to build a model that can predict the condition of the tool (worn/unworn) when I provide one of the experiment runs.

How should I approach this problem? I don't quite get the flow to build the predicting model. I was thinking using Random Forest Classification to do the model. I don't understand how to make binary predictions based on a group of time-series data