outhimar_64 Close regression

Introduction

The outhimar_64-Close-regression is a baseline model designed to perform regression analysis on tabular data using Scikit-learn. It specifically applies regression to predict the 'Close' value in financial datasets, achieving high accuracy with an R² score of 0.999858.

Architecture

The model utilizes a Scikit-learn pipeline comprising two main components:

1. EasyPreprocessor: This component preprocesses the data by identifying and processing various data types such as continuous, date, and categorical features.
2. Ridge Regression: The model is built using Ridge regression with an alpha parameter of 10, providing regularization to avoid overfitting.

Training

The model was trained using the DABL library as a baseline, demonstrating strong performance metrics with minimal mean squared error. The training process involved experimenting with different models and configurations, documented in logs.txt.

Guide: Running Locally

To run the model locally, follow these steps:

1. Install Dependencies: Ensure Python and Scikit-learn are installed. Additionally, install the DABL library if not already present.

   pip install scikit-learn dabl
   

2. Download the Model: Clone or download the model files from the Hugging Face repository.
3. Load and Preprocess Data: Use the EasyPreprocessor to prepare your dataset similar to the training data format.
4. Run the Model: Execute the Ridge regression with the preprocessed data to get predictions.

For more computationally intensive tasks, consider using cloud GPU services like AWS EC2 with GPU, Google Cloud Platform, or Azure.

License

This model is released under the Apache-2.0 license, allowing for open use, modification, and distribution.