leonardkarsunky Goto Github PK

ai_project_heart_disease

copd_analysis

The aim of this analysis is to cluster/segment COPD patients into distinct subpopulations using unsupervised learning techniques. The dataset includes 101 (imaginary) patients and 22 features (both numerical and categorical), including information on their characteristics, disease severity, and co-morbidities.

css-exercises

deeplearning_2d_classification

Built a deep learning framework using neural networks ”from scratch”, i.e. without using autograd or the neural network modules already implemented in PyTorch. Tested using 2D point classification.

deeplearning_digit_comparison

Built pipeline to test different architectures of deep neural network to compare two digits visible in a two-channel image. Assessed importance of weight sharing and auxiliary loss.

lab-simulation-of-mice

C++ program that simulates the external dynamic environment of a laboratory studying mice, including the mice's boxes, the mice's livers, the livers's cells (including cancer cells), and the cells's diffused substances, all in real time using a custom UI.

landing-page

leonardkarsunky

Config files for my GitHub profile.

ml-zlf-project2

Retinal fundus image processing pipeline for image registration using unsupervised learning algorithms (SIFT, AKAZE, VoxelMorph) implemented with OpenCV.

networkcode---final

odin-recipes

pathml

A Python library for deep learning on whole-slide images

pathml-tutorial

A tutorial demonstration of pathml on an example dataset for the manuscript "PathML: A unified framework for WSI analysis with deep learning" by Berman et al.