👇

---

# 📜 Competition Rules

Welcome to the **SolarSpectra Challenge** — a 24-hour datathon to predict solar panel efficiency using machine learning.
To keep the competition fair and collaborative, please follow the rules below.

---

## 👥 Team Rules

* Teams can have **2–4 members**.
* Each team must make submissions **under a single Kaggle account**.
* Collaboration **within a team** is encouraged, but **between teams** is not allowed.

---

## 📂 Data Rules

* You may use **only the data provided** in the competition dataset.
* **External data** or pretrained models are **not allowed**, unless explicitly approved by organizers.
* You may perform any **data cleaning, feature engineering, or modeling** on the provided data.
* The dataset contains both **useful and irrelevant features** — part of the challenge is identifying which ones matter.

---

## ⚙️ Submission Rules

* You may submit up to **5 times per day**.
* Your **best public leaderboard** score will appear during the event.
* The **private leaderboard** (based on 70% of test data) determines the final winners.
* The submission format must exactly match the sample submission:

  ```
  id,efficiency
  0,10.542
  1,11.837
  2,9.428
  ```
* Predictions must be **numerical** and correspond to **all rows** in the test set.

---

## 🔒 Code & Fair Play

* Code must be **original** and created during the competition window.
* Use of **AI assistants (e.g., ChatGPT, Gemini, Copilot)** for learning or debugging is allowed,
  but **auto-generation of full model code or feature logic** is discouraged — this is a data understanding challenge.
* You may share **insights** after the competition, but not during the event.
* Plagiarism, code leaks, or unauthorized sharing of solutions will result in disqualification.

---

## 🏁 Evaluation & Winners

* Metric: **Root Mean Squared Error (RMSE)**
* Public leaderboard: **30% test data**
* Private leaderboard: **70% test data** (final ranking)
* In case of a tie, the team with **earlier final submission** will be ranked higher.

---

## 🧠 Organizer’s Notes

This competition is designed to test:

* **Scientific reasoning** — understanding what drives solar efficiency.
* **Data intuition** — feature selection, handling noise, and dealing with missing data.
* **ML modeling** — regression techniques, ensembling, and validation.

Think critically, collaborate wisely, and build models that **shine under real-world conditions** 🌞

---