AI Fuzzy Logic EOR Predictor

Budget: $10 - $30

I am building an Enhanced-Oil-Recovery screening tool that estimates incremental oil gain through a fuzzy-logic model written in Python. Here is what I need from you:

The core model • Construct a fuzzy inference system (scikit-fuzzy or equivalent) that learns from both simulated and real data. • Simulated data must come from uniformly distributed random arrays generated between specified high-low bounds. • Use the following inputs throughout the workflow: Temperature, Porosity, API, Viscosity, Oil saturation, Depth, and Permeability. Pressure can remain optional for future expansion but is not required right now. • Feed published field data into the training phase so the network benefits from real-world behaviour, then reserve a slice for blind testing.

Analytics and visuals • Produce learning-curve plots that track loss vs. epoch and overlay training / validation histories. • Generate clear prediction vs. actual graphs for every parameter above, along with aggregate error metrics (RMSE, MAE).

Interactive interface • A lightweight GUI (Tkinter, PyQt, or similar) must allow users to enter each parameter manually and receive the predicted incremental oil gain instantly. File upload is not needed at this stage, so focus on clean manual entry fields, a “Run Prediction” button, and a results panel that embeds the graphs created earlier.

Acceptance criteria

1. Python source code, fully commented and runnable end-to-end.
2. Reproducible data-simulation routine with seed control.
3. At least three example learning-curve screenshots and parameter-specific prediction plots.
4. Executable GUI that returns consistent results with the command-line version.

If you are comfortable with fuzzy logic, EOR concepts, and quick desktop GUI builds, let’s talk timelines and milestones.