search

nagdevAmruthnath / Predictive-Maintenance

A notebook tutorial series for performing predictive maintenance using machine learning
MIT License
144 stars 74 forks source link

Project Flow #1

Closed vidhithakkar12 closed 4 years ago

vidhithakkar12 commented 4 years ago

Hello Nagdev,

Thank you for sharing your project. I would like to understand the project flow and description of the data that you have used. Looking forward to a quick response.

Thanks, Vidhi

nagdevAmruthnath commented 4 years ago

@vidhithakkar12 Please see my notebooks for the project flow.

  1. Load the data
  2. Summarize the data and extract features
  3. Preprocess data
  4. Understanding features
  5. Set up experiments with different model, various sample sizes and hyper parameters
  6. Build models for anomaly detection or fault classification
  7. Test the models
  8. Verify and validate the consistency of the results

The data used is vibration data with 3 axes (XYZ) with sample rate or 100Hz or 100 samples per second.