Kaggle

A collection of Kaggle solutions.

Real Estate Price Prediction

• In this competition your task will be to predict the price of flats in test.csv.
• Model: Gradient Boosting

Credit Scoring

• Your task will be to build a scoring model for the bank's secondary customers that would predict the probability of a customer default.
• Model: Logistic Regression

Restaurant Rating prediction

• You need to create a model that will predict the rating of the restaurant according to TripAdvidor.
• Model: Random Forest

Prediction of user churn

• You will have to learn how to simulate the outflow of telecom company customers.
• Model: Stacking (Logistic Regression, CatBoost, KNN, Random Forest)

Prediction the hotel rating on Booking

• One of the company's problems is dishonest hotels that wind up rating themselves. One of the ways to find such hotels is to build a model that predicts the rating of the hotel.
• Model: Random Forest

Prediction the car price

• A company sells used cars, and at some point a manager comes to you with a proposal to build a model to predict the adequate cost of the car based on its characteristics from the ad.
• Model: LightAutoML

Journey to Springfield

• Train a classifier based on Convolutional Networks to learn how to distinguish all residents of Springfield..
• Model: CNN

New York City Taxi Trip Duration

• Build a model that predicts the total ride duration of taxi trips in New York City
• Model: RandomForest

Capture recongnition

• A problem from the field of computer vision is proposed – recognition of the letters of the English alphabet on distorted and noisy images
• Model:

Text classification

• The task is to predict the genre of the film (for example, comedy, drama, etc.) from a brief description of the plot.
• Model: BERT

Image segmentation

• The dataset proposes to check skills in mastering of determination of lips' position on photos.
• Model: Linknet

Text classification

• The objective of this task is to detect hate speech in tweets.
• Model: CNN

Sentiment analysis

• The objective of this task is to predict of sentiment messages from tweets.
• Model: RNN (GRU/LSTM)