Flash Point Temperature Prediction using MLP

This project predicts the flash point temperature of chemical compounds using Joback subgroups as input features. A multi-layer perceptron (MLP) model is trained using TensorFlow for the task. The model maps the number of subgroups in a molecule to its corresponding flash point temperature.

Model Details

• Input features: Joback subgroups (number of subgroups for each compound)
• Labels: Flash Point Temperature (K)
• Framework: TensorFlow
• Evaluation Metrics: MAE, MSE, R²

Repository Contents

• Flash_Point_Temperature.ipynb: Jupyter notebook containing the model training and evaluation.
• model/: Directory containing saved models.
• data/: Sample dataset with chemical compound SMILES and flash point temperatures.
• requirements.txt: Dependencies for running the project.

Usage

1. Clone the repository:

   git clone https://github.com/yourusername/flash-point-prediction.git

2. Run the main.py.
3. write the SMILES of the chemical compound to predict its Flash Point Temperature.