Scribd
Upload
20 views

Final_Project

The final project for the Master in Big Data Analytics involves forecasting share value and share volume for 8 brands over the next 52 weeks using both statistical and machine learning methods. Teams of up to 3 students must submit a notebook, a CSV file with predictions, and a compiled report by March 20th, 2025. Evaluation criteria include prediction accuracy, predictive methodology, model explanations, and originality in documentation.

Uploaded by

Ramón García
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
20 views

Final_Project

The final project for the Master in Big Data Analytics involves forecasting share value and share volume for 8 brands over the next 52 weeks using both statistical and machine learning methods. Teams of up to 3 students must submit a notebook, a CSV file with predictions, and a compiled report by March 20th, 2025. Evaluation criteria include prediction accuracy, predictive methodology, model explanations, and originality in documentation.

Uploaded by

Ramón García
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 6

Time series analysis and forecasting

Master in Big Data Analytics

Final Project: Advanced forecasting tools

Deadline: March 20th, 2025, at 13:00 (no delays are allowed)


Upload to Aula Global: notebook and csv

Master in Big Data Analytics Time series analysis and forecasting Final Project 1/6
General Objectives

• The goal is to predict the 2 target variables (Share value and Share volume) for 8 brands (i.e.
8x2 time series) on a weekly basis.

• Work in teams of max 3 students. You need to register your team here: Team Registration Sheet.

• For each time series, you will have 2 regressors: Price and DP, along with the weekly date.

• Your task is to forecast the 16 time series for the next 52 weeks.

• All models should include both statistical and machine learning tools, as well as methods for
generating prediction intervals (only for the notebook).

Master in Big Data Analytics Time series analysis and forecasting Final Project 2/6
Competition Setup and Data

• The training dataset (provided on Aula Global) consists of the historical time series for all brands,
with the predictors (Price and DP) and the targets (Share value and Share volume).

• Your goal is to forecast the next 52 weeks for each brand and each target variable.

• Forecasts should be produced on a weekly basis.

Master in Big Data Analytics Time series analysis and forecasting Final Project 3/6
Performance Evaluation
• Prediction Accuracy (30%):
• Your predictions for the hidden test set (last 52 weeks) will be evaluated using Mean Absolute Error
(MAE).
• Teams will be ranked based on MAE.

• Predictive Methodology (30%):


• Use advanced statistical and machine learning tools, including automatic tools, dynamic regression,
multivariate time-series, and ML
• Justify your model choices and validation procedures.

• Model Explanations (20%):


• Identify key drivers for Share value and Share volume, with emphasis on the effects of Price and DP.

• Originality and Documentation (20%):


• Innovative approaches, feature engineering, and thorough documentation will be rewarded.

Master in Big Data Analytics Time series analysis and forecasting Final Project 4/6
Submission Format for Predictions (CSV)

• The prediction file must include the following columns: Brand, Date Monday, Share value pred,
Share volume pred.

• The file should contain 52 rows for each brand (a total of 8x52 rows).

• Example: Brand,Date Monday,Share value pred,Share volume pred

Brand1,2024-02-04,15.5,12.8
Brand1,2024-02-11,14.7,11.1

Master in Big Data Analytics Time series analysis and forecasting Final Project 5/6
Submission Guidelines

• The training dataset is available on Aula Global.

• Submit the following:


1 Notebook (R Markdown or Jupyter Notebook): Include all code, visualizations, and analysis.
2 CSV File (Predictions): Your forecasts for the test set in the required format, named as
TeamName.csv.

3 Compiled Report (HTML or PDF): An exported version of your notebook.

• Name your notebook file as: TeamName LastName1 LastName2.Rmd or


TeamName LastName1 LastName2.ipynb.

• Cite all external sources or code snippets used in your analysis.

Deadline: March 20th, 2025, at 13:00

Master in Big Data Analytics Time series analysis and forecasting Final Project 6/6

You might also like