MLops_G2JN: Automatic Pipeline for fixing uncertainties

Introduction

We present in this repo our pipeline spotting uncertaincies in a given dataset, and then applying automatic solution on inspected epistemic and aleatoric uncertainties.

• For inspecting uncertainties we use MACEst.
• For data generation we use SDV.
• For outliers removal we use PyOD.

Running the pipeline

In order to activate our pipeline you can do the follwoing steps:

1. Press Start_G2JN.ipynb
2. Run the first cell:

%%capture
!pip install macest
!pip install sdv
!pip install pyod
import os
os.kill(os.getpid(), 9)

The kernel will crash and restarts after finishing the installments - thats ok!
3. Run the following cell and watch output:

!git clone https://github.com/gilzeevi25/MLops_G2JN 
%cd MLops_G2JN
!python main.py 

Client's reports

Please click here to inspect:

• Midterm report
• Final report

Thanks for visiting