Time Series Project Tutorial

• Understand a new dataset.
• Analyze the time series and study its characteristics.
• Train a model to predict the amount of water in different areas.

🌱 How to start this project

Follow the instructions below:

1. Create a new repository based on machine learning project by clicking here.
2. Open the newly created repository in Codespace using the Codespace button extension.
3. Once the Codespace VSCode has finished opening, start your project by following the instructions below.

🚛 How to deliver this project

Once you have finished solving the exercises, be sure to commit your changes, push them to your repository, and go to 4Geeks.com to upload the repository link.

📝 Instructions

Water detection system

This project is going to be done at Kaggle. Kaggle is known for organizing data science competitions in which individuals and teams can compete to create the best model on a variety of tasks.

Specifically, we will explore a competition that has now ended, which rewarded the top performers with $25,000, distributed to the top four solutions.

Step 1: Register or log in to Kaggle

The first step to participate in a competition is to have a Kaggle account. Registration is very simple and versatile.

Step 2: Participate in the competition

Although it is closed, you can still upload your solutions and compare yourself with the rest of the competitors in the ranking. By clicking here, you can access the competition, where the purpose is to predict the amount of water.

Read the information about the problem description, evaluation metrics, timeline and prizes. Once you have an overview of the competition, prepare the notebook with the code development and upload it. Compare yourself with the other developers.

Follow these tips:

• Look at the solutions proposed by other developers. Some use time series, others use supervised algorithms such as linear regression, decision trees, etc. Investigate which model might work best for this use case and explore all options.
• Whatever model you choose, be sure to process and explore the data as studied in past modules.
• Trial and error! Don't pretend to get the best model the first time, keep trying.

Note: We also incorporated the solution samples on ./solution.ipynb that we strongly suggest you only use if you are stuck for more than 30 min or if you have already finished and want to compare it with your approach.