https://github.com/bnvulpe/regression-and-time-series

This work centers on assessing and comparing predictive models for regression and time series prediction using specific datasets, with the goal of selecting the most effective methodology for unseen test data.
https://github.com/bnvulpe/regression-and-time-series

colab data-analysis data-analysis-python data-science data-visualization forecasting jupyter-notebook machine-learning model-evaluation predictive-modeling python regression sarima sarimax time-series-analysis time-series-analysis-and-forecasting

Last synced: about 15 hours ago
JSON representation

This work centers on assessing and comparing predictive models for regression and time series prediction using specific datasets, with the goal of selecting the most effective methodology for unseen test data.

• Host: GitHub
• URL: https://github.com/bnvulpe/regression-and-time-series
• Owner: bnvulpe
• License: mit
• Created: 2024-01-21T16:05:04.000Z (10 months ago)
• Default Branch: main
• Last Pushed: 2024-06-22T11:48:15.000Z (5 months ago)
• Last Synced: 2024-06-22T20:31:43.853Z (5 months ago)
• Topics: colab, data-analysis, data-analysis-python, data-science, data-visualization, forecasting, jupyter-notebook, machine-learning, model-evaluation, predictive-modeling, python, regression, sarima, sarimax, time-series-analysis, time-series-analysis-and-forecasting
• Language: Jupyter Notebook
• Homepage:
• Size: 10.9 MB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
Unveiling Optimal Predictive Models for Regression and Time Series Analysis.

 ---------------------------------------

The main focus of this work lies in the evaluation and comparison of the predictive ability of various models within the specific context of the provided datasets. The ultimate purpose is the selection of the methodology that demonstrates optimal performance when faced with unseen data in the test set. 

To achieve this goal, a detailed description of the work process will be carried out, supported by the use of graphics that provide a clear visualization of the decisions made and the results obtained. The aim is not only to present the final results but also to provide a deep understanding of the methodological choices, adopted approaches, and any key considerations influencing the interpretation of the results. 

We will address two different problems, each associated with a specific dataset. We will use the datasets `train_ap1_mcp_23_24_train.csv` and `test_ap1_mcp_23_24_test.csv` to perform the regression task, specifically predicting the popularity of songs. Additionally, we will employ the dataset `nacimientos_2016_2021.csv` to make time series predictions for births. 

The organization of the work will revolve around these datasets, with the first two sections dedicated to the regression problem and the last one focused on the time series.

 ---------------------------------------

All code, CSV files, and any other necessary files for executing the code are available in the 'code' folder of this repository. The code has been modified for ease of execution in the Google Colab environment. A comprehensive report of the work undertaken is available in Spanish (considered to be translated to English).