log well prediction

Welcome to the log well prediction repository! This project focuses on predicting specific targets using machine learning models. Below, you'll find a comprehensive guide to get started with the project, including data preparation, model training, and result evaluation.

Table of Contents

1. Introduction
2. Setup
3. Data Preparation
4. Model Training
5. Results
6. Contributing
7. License

Introduction

This project leverages Transformer-KAN techniques to predict log well from a given dataset. The main workflow involves data preparation, model training, and evaluating the model's performance.

Setup

Before you begin, ensure you have the necessary dependencies installed. You can do this by running:

pip install -r requirements.txt

Data Preparation

First, you need to prepare your data by running the data_pre.py script. This script processes your dataset and saves it in the specified directory. If you need to change the prediction target, you can modify the following code in the tool_for_pre module:

parser.add_argument('--input_directory', type=str, default=r'data_save\54口井的数据集', help='输入地址')

Please go to for complete data https://www.kaggle.com/datasets/charzhang/well-log-data/data download

Your data should be stored in the data_save directory. For testing purposes, you can use the dataset provided in the 数据读取的案例数据 file.

To run the data preparation script, use the following command:

python data_pre.py

Model Training

Next, train your model using the train.py script. You can execute the training process from the command line with customizable parameters. Here is an example command:

python train.py --model_name Transformer_KAN --hidden_size 32 --num_layers 4 --num_heads 4 --num_epochs 200 --learning_rate 0.001 --input_directory data_save\数据读取的案例数据 --input_size 5 --batch_size 32 --sequence_length 20 --predict_target "DEN"

Parameters:

• --model_name: The name of the model to be used (e.g., Transformer_KAN).
• --hidden_size: Size of the hidden layers.
• --num_layers: Number of layers in the model.
• --num_heads: Number of attention heads.
• --num_epochs: Number of training epochs.
• --learning_rate: Learning rate for the optimizer.
• --input_directory: Directory where the input data is stored.
• --input_size: Size of the input features.
• --batch_size: Batch size for training.
• --sequence_length: Length of the input sequences.
• --predict_target: The target variable to predict (e.g., "DEN").

Results

After training, the model outputs the results, which are stored in the model_save directory. You can check the prediction results in this directory to evaluate the performance of your model.

Contributing

Contributions are welcome! Please feel free to submit a Pull Request or open an Issue to improve the project.

License

This project is licensed under the EIAS License. See the LICENSE file for more details.

---

Thank you for using log well prediction! If you have any questions or need further assistance, please feel free to contact us.