Data Science Projects:

• House rent price predictor. The project aims to predict apartment rent prices based on inputparameters using a dataset of Ukrainian rental announcements. Classical regression machine learning algorithms, including Linear Regression, Random Forests, Gradient Boosting, and XGBoost, were applied. The final model achieved an R2 score of 0.82 on the validation dataset. A simple web app was also created: https://ukrainianhouserentpricepredictor.streamlit.app.
  Code: github.com/Vladfsociety/house_rent_price_predictor
• Custom neural network. Mini-library that implements a simple version of a feedforward neural network (FNN) and convolutional neural network (CNN) from scratch using Python and PyTorch.
  Code: github.com/Vladfsociety/vladk-neural-network
• Animal class predictor. Project contains implementations of various convolutional Neural Networks for animal classification task using PyTorch and TensorFlow. Best classifier (VGG16 pretrained on ImageNet PyTorch model) achieved 95% accuracy on validation dataset.
  Code: github.com/Vladfsociety/animal_classification