https://github.com/gblssroman/vk_internship

🏢 EDA (incl. KNN) & CatBoost & Optuna for predicting the most accurate scores for the business objects success.
https://github.com/gblssroman/vk_internship

catboost geolocation knn machine-learning vk

Last synced: 5 days ago
JSON representation

🏢 EDA (incl. KNN) & CatBoost & Optuna for predicting the most accurate scores for the business objects success.

• Host: GitHub
• URL: https://github.com/gblssroman/vk_internship
• Owner: gblssroman
• Created: 2024-03-20T19:13:23.000Z (9 months ago)
• Default Branch: master
• Last Pushed: 2024-03-20T23:38:38.000Z (9 months ago)
• Last Synced: 2024-10-31T06:25:21.945Z (about 2 months ago)
• Topics: catboost, geolocation, knn, machine-learning, vk
• Language: Jupyter Notebook
• Homepage:
• Size: 32.6 MB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# 🏢 VK Internship March '24

EDA (incl. KNN) & CatBoost & Optuna for giving the most accurate scores for the object success prediction task.

Repo for VK Internship Data Science task.

```.zip``` archive included.

**Instruction:**

1. Install ```requirements.txt``` via ```pip install -r requirements.txt``` (no junk :))

1. Launch ```generate_submission.py```

2. Get final ```submission.csv``` (it has already been generated in the folder ```output``` for fast reference).

Other files description:

* ```classifier.cbm``` - Trained CatBoost model (regressor, not classifier)

* ```do_eda.py``` - Script used in ```generate_submission.py``` for given datasets preparation

* ```datasets``` - Folder containing datasets

* ```cols_to-drop.pkl``` - Columns to-be-dropped causing multicollinearity (dict).

---

**Инструкция:**

1. Установите ``requirements.txt`` через ``pip install -r requirements.txt``

1. Запустите ``generate_submission.py``

2. Получите окончательный ``submission.csv`` (он уже сгенерирован в папке ``output`` для быстрого референса).

Описание других файлов:

* ```classifier.cbm``` - Обученный регрессор CatBoost

* ``do_eda.py`` - скрипт, используемый в ``generate_submission.py`` для подготовки датасетов и фичей

* ``datasets`` - Папка, содержащая наборы данных

* ``cols_to-drop.pkl`` - Столбцы, подлежащие удалению, вызывающие мультиколлинеарность и по факту не дающие полезной информации.