The purpose of this repository is to share the code and models developed in during the Predictive Analysis internship project done by Rishit Arora, Avani Yerva and Max Rontal
- Clone this repository and install the necessary Python packages
- Make sure to have the data_preprocess.py file in the same directory as the other .ipynb files as they are dependent on the python file
Many functions and lines of code have comments that explain what they do
- basic_graphs.ipynb is used to visualize basic FFT graphs
- BEA_with_STFFT_and_SK.ipynb is used to perform bearing envelope analysis with information gathered through spectral kurtosis of Short Time Fast Fourier Transform on the microphone data
- data_preprocess.py is used to preprocess the microphone data such as performing FFT and cutting off certain frequencies
- ml_analysis_cnn_inside_oven.ipynb is used to create models
- model_testing.ipynb is used to test models that are created
- convert_model.ipynb is used to convert the model to ONNX format for the Microsoft.ML.OnnxRuntime library to use in .NET Desktop Applications (Deployment)