Data science notebooks

This repository comprises a compilation of my solutions and experiments in machine learning problems. Jupyter notebooks encompass research on open-source datasets and solutions to Kaggle competitions. Here, you can explore a diverse range of approaches and techniques for data analysis, along with the application of machine and deep learning methods.

Kaggle

• Simpsons classification, simpsons.ipynb (DLS assignment)

MIPT ML courses

Generic

• Wine dataset, wine.ipynb
• Regression task with a raw dataset, raw_dataset_regression.ipynb

Computer vision

• Breast cancer cells segmentation, cancer_cells_segmentation TODO: add classification metrics such as Precision, Recall, etc.
• Panorama stitching

Advanced

Repository

• Analysis of CNN on the EMNIST-letters dataset
• Prediction of POS tags for tokens in the NERUS dataset using LSTM
• Analysis of an autoencoder in the sentence reconstruction task for the Twitter dataset
• Generation of image annotations for the COCO dataset

Deep Learning School

• PH2 Dataset images, ph2_semantic_segmentation.ipynb