For budding Data Science and AI engineers, it is very useful to practice, demonstrate and participate in competitions. Kaggle solves this problem. This article provides the most practical steps to get started with Kaggle.

Kaggle-Data Science and AI is a community of engineers. Although it originally started as a racing platformer, many sections were added later. As a result, Kaggle has become a platform with many tools for becoming an expert from scratch: from mini-courses to learn ML to research competitions.

To start classes (or competitions) on Kaggle, we first register on the site. The platform is designed like a game, as soon as you register, your status will be below:

Novice (New).

As you do certain activities, your status (level) will increase, and you will become a Contributor (Participant).

• 1 script or notebook should run
• Must submit 1 in any competition
• 1 must comment
• 1 upvote (something like like)

Expert

Master

Grandmaster

Competitions

1. Featured: These contests are usually published by large companies, organizations and even governments. Their cash prizes are much larger than those offered in other categories.
2. Research: These are research-themed competitions. Little or no prize money.
3. Start: These do not include any premiums. They are generally competitions created for educational purposes. At the end of this section, you’ll find a sample contest labeled “Getting Started.” You will not only see the example, but also a guide on how to use the notebook and how to present the results, among other relevant steps.
4. Playground: These are suitable competitions for those who want to gain some experience and continue to improve their skills. Prizes are usually Kaggle merchandise (such as t-shirts and stickers). These contests are often fun and games in nature.
5. InClass: These are competitions usually run by universities and their participants are machine learning students. Their goal is to engage and inspire these students.
6. Analytics: These are data analysis competitions.
7. Simulations: What differentiates these from traditional supervised machine learning challenges on Kaggle are the types of competitions that are reinforcement learning tasks. Competitors build models and allow their models to compete in a simulated environment

Datasets
One of the most useful sections on Kaggle. You can find all kinds of datasets. You can also create a Dataset yourself and upload it to Kaggle. You can see a list of the most popular Datasets from this link. Note that there are many datasets related to Azerbaijan.

Codes
An online editor that lets you create a Jupyter Notebook or simple script in Python and R. You simply connect data and work in the browser without installing libraries. The code posted is very diverse: from EDA (known as detailed analysis) of racing problems to simple techniques to help you optimize your programs.

Discussions
The discussion is divided into several main sections:

General Kaggle news, announcements, and contest discussions
Getting Started-Similar to above, for beginners.
Product Feedback – Go here for issues on Kaggle.
Questions&Answers-Support from other Data Scientists
Discussions about lessons in the Learn-Courses section