Skip to content

Latest commit

 

History

History
83 lines (47 loc) · 3.42 KB

README.md

File metadata and controls

83 lines (47 loc) · 3.42 KB

Band Gap Prediction of Inorganic Semiconductors

Overview

The Band Gap Prediction project aims to predict the band gap of various materials using machine learning techniques. The band gap is a crucial property that determines whether a material is an insulator, semiconductor, or conductor. This prediction can be valuable in material science, electronics, and renewable energy research.

Background

The electronic band gap is the energy difference between the highest occupied energy state "valence band" and the lowest unoccupied energy state "conduction band" in a material. It plays a fundamental role in the electrical and optical properties of materials. This project aims to build a predictive model to estimate the band gap based on material characteristics.

Dataset

The Dataset has been retrieved with the help of matminer huge repository. expt_gap Experimental band gap of 6354 inorganic semiconductors.

the Paper and the dataset used:

Refrence
Dataset json.gz format

Installation

To set up the project environment, follow these steps:

  1. Clone the repository:

    git clone https://github.com/achraf110/Band_gap_prediction.git
cd Band_gap_prediction

  2. Install miniconda

  3. Create a virtual environment (optional but recommended):

    conda create -n myenv 
conda activate myenv

    replace myenv with the name of your env of choice

  4. Install the required dependencies :

    pip install -r requirement.txt

Feature Engineering with Matminer

With the help of .featurizers package from Matminer library, we exacted a wide range of features to properly describe our list of Inorganic Semiconductors. Refer to the code to see more.

Model Training and Analysis

We employ two machine learning algorithms for band gap prediction:

  1. Linear Regression: We utilize linear regression to establish a baseline for band gap prediction and to understand the general trend between features and band gaps. We will visualize the results and evaluate the model's performance.

  2. Random Forest Regression: The random forest algorithm provides a more sophisticated and robust approach for predicting band gaps. We investigate the feature importance provided by the random forest model to identify the most influential features affecting the predictions.

Results and Visualization

We present the results of our band gap predictions in an interactive manner. The visualizations will showcase the relationship between material features and band gaps, emphasizing the most significant features discovered through feature importance analysis.

giphy.mp4

License

GNU General Public License v3.0

Changelog

  • work in progress ...

Acknowledgments

We extend our gratitude to the Matminer and Magpie developers for their invaluable contributions to the materials informatics community. We also acknowledge the datasets used in this project and the research papers that inspired our work.

Contact

For any questions, feedback, or collaboration opportunities, feel free to send me an email.

[email protected]

Chahbi Achraf.