By Cognivio Team
This repository contains our submission for the Data Mining Competition held during the Intelecta Cup Telkom University 2025 in the qualifying round.
The main objective of this case study is to forecast future climate conditions based on multivariate time-series data.
We experimented with several ensemble-based approaches consisting of multiple weak learners.
Below is a summary of our simplified experiment results:
| No | Model | MAPE (%) |
|---|---|---|
| 1. | Linear Regression, Naive Drift, Naive Seasonal | 146.533 |
| 2. | LightGBM, Naive Drift, Naive Seasonal | 151.973 |
| 3. | LightGBM, XGBModel, Naive Drift, Naive Seasonal | 129.088 |
| 4. | XGBModel, Linear Regression, Naive Drift, Naive Seasonal | 142.532 |
Among these configurations, the LightGBM + XGBModel + Naive Drift + Naive Seasonal ensemble (Experiment 3) achieved the best performance.
You can view the corresponding notebook here:
notebooks/final/cognivio-final-notebook.ipynb
We also prepared a paper explaining our full methodology and findings.
You can find it in the paper directory under the filename:
PENYISIHAN_MAKALAH_INTELECTA2025_DATAMINING_COGNIVIO.pdf
The paper is fully written in Bahasa Indonesia.
In the qualifying round, our team ranked 30th out of 90 participants,
with a final score of 58.152 / 100.
You can view the complete scoreboard using the link below:
Competition Scoreboard