DominikaBakalarz

AdaBound

An optimizer that trains as fast as Adam and as good as SGD.

Applied-and-Computational-Statistics

CNN-with-Fashion-MNIST-dataset

This is my first implementation of a network. I have used a 2 layered CNN with maxpooling and Xavier weights.

ConfocalMicroscopy

Part of the Origami BioBandage project. The code is supposed to read the confocal microscopy pictures and recognize alive and dead cells. Files lodepng.h and lodepng.cpp are required extra libraries. File stemscan.cpp contains a code that I have written myself. The code is supposed to read the pictures from the folder input-pictures (one picture at a time) and then it recognises alive and dead cells (alive cells are green, dead are red and black colour represents the background, I’ve calculated that cell of this kind (stem cell) is 16 pixels wide on average) and marks them with white and blue dots, and returns modified picture.

Dissertation-2019

Review of deep learning methods for lung cancer detection using Computed Tomography images.

Origami_BioBandage

Origami BioBandage - personalised therapy for bone fractures based on mathematical modelling of bioimplants.

pacbayes-opt

Optimizing PAC-Bayes bounds for Stochastic Neural Networks with Gaussian weights

oxai

Oxford Artificial Intelligence Society

DominikaBakalarz/AdaBound

An optimizer that trains as fast as Adam and as good as SGD.

DominikaBakalarz/Applied-and-Computational-Statistics

DominikaBakalarz/CNN-with-Fashion-MNIST-dataset

This is my first implementation of a network. I have used a 2 layered CNN with maxpooling and Xavier weights.

DominikaBakalarz/ConfocalMicroscopy

Part of the Origami BioBandage project. The code is supposed to read the confocal microscopy pictures and recognize alive and dead cells. Files lodepng.h and lodepng.cpp are required extra libraries. File stemscan.cpp contains a code that I have written myself. The code is supposed to read the pictures from the folder input-pictures (one picture at a time) and then it recognises alive and dead cells (alive cells are green, dead are red and black colour represents the background, I’ve calculated that cell of this kind (stem cell) is 16 pixels wide on average) and marks them with white and blue dots, and returns modified picture.

DominikaBakalarz/Dissertation-2019

Review of deep learning methods for lung cancer detection using Computed Tomography images.

DominikaBakalarz/Origami_BioBandage

Origami BioBandage - personalised therapy for bone fractures based on mathematical modelling of bioimplants.

DominikaBakalarz/pacbayes-opt

Optimizing PAC-Bayes bounds for Stochastic Neural Networks with Gaussian weights