Machine Learning Assisted Prediction for Quantum Well Photoluminescence Growth

A machine learning (ML) model used to look at photoluminescence (PL) data from our InGaAs quantum wells. This model will provide insight on future quantum wells grown by using a supervised learning agent.

Introduction

This is a small project that extracts the peak photoluminescense (PL) and intensity out of a previously measured Quantum Well sample. Then, using a machine learning model, you can analyze this with other relevant input parameters to predict peak PL measurements for future Quantum Wells grown.

Dependencies

This project uses three main libraries: RandomForestRegressor, GradientBoostingRegressor, and MultiOutputRegressor.

For more information, please visit their respective repositories and documentation.

• RandomForestRegressor
• GradientBoostingRegressor
• MultiOutputRegressor

Virtual Environment Setup

1. Install necessary packages:

   pip install pandas
   pip install numpy
   pip install scikit-learn
   pip install matplotlib
   pip install openpyxl

How to Use the Code

1. Create the Raw PL Spectra Folder

   • Inside your desired directory (e.g. PL_ML_Project), create a new folder named PL_Spectra_Raw
   • Place all your individual raw PL spectrum files (e.g., G25-017-center.csv, etc.) into this PL_Spectra_Raw folder.
   • The filename should exactly match the QW_Sample ID you'll use in your quantum_well_data.xlsx file.
   • Internal Format: Each file should have comment lines starting with #, followed by a header line (e.g., lambda [nm] intensity [a.u.]), and then tab-separated numerical data.

2. Create your quantum_well_data.xlsx file

   • Populate this file with your actual data inputs. For example:
     • QW_Sample
     • In_Fraction
     • WellWidth_nm
     • etc.

3. Configure and Run the Peak Extraction Code

   python pl_peak_extractor.py

   • Verify a file named extracted_pl_peaks.csv (or .xlsx) appears in your PL_ML_Project directory.

4. Configure and Run the Machine Learning Code

   • Ensure FEATURE_COLUMNS = ['GrowthTemp_Pyro', 'PL_Correction_Factor', etc.] match your exact varying column names.
   • Adjust IDEAL_PL_WAVELENGTH and IDEAL_PL_INTENSITY to your specific targets.
   • Adjust example prediciton values in the --- Example Recipes --- section.
   • Save the file and run:

   python main_ml_script.py

Example Plots

Plot of Predicted vs Actual Peak Wavelength & Peak Intensity

Example of Photoluminsence Plot for Training