Filtyp
Artificial Neural Networks (ANNs) are widely used information processing algorithms based roughly on biological neural networks. These networks can be trained to find complex patterns in datasets and to produce certain output signals given a set of input signals. A key element of ANNs are their so-called activation functions, which control the signal strengths between the artificial neurons in a n
Artificial neural networks have been used to solve different problems, one being survival analysis of medical data. For survival analysis, the main interest is often how samples become sorted by their outputs, which makes survival analysis a rank based problem. A rank based error function lacks a gradient, which makes gradient descent based training difficult. A genetic training algorithm, based o
Convolutional Neural Networks (CNNs) and pre-trained word embeddings have revolutionized the field of Natural Language Processing (NLP) during the last years. In this project, CNNs are used on top of the Word2Vec word representation for a sentence classification task on medical research articles. Both individual networks for each category as well as a combined classification network are optimized
In the field of computational biology, fluorescence microscopy images often constitute the input source of information. The process of binarization of raw images to delineate interesting objects requires image segmentation into signal and background pixels. Several methods to perform image segmentation exist, the Otsu method being a popular unsupervised example. The Otsu method's lack of pro
The beneficial effects of HLA matching at the HLA-A, -B and -DR loci on recipient short- and long-term survival in heart transplants have been established in a number of studies. The objective of the present study was to evaluate the importance of individual HLA alleles at the HLA-A, -B and -DR loci as risk factors for recipient death within one year after transplant by using artificial neural net
Previously, Berg and Purcell found an expression for the diffusion current of particles into a set of receptors in their paper Physics of Chemoreception. These receptors were assumed to be uniformly distributed on a spherical cell's surface and the receptors where idealized as circular patches. For their model, the particles' diffusion constant was the same for all points outside of the ce
This thesis examines the state-of-the-art 2D super-resolution technique alternating descent conditional gradient (ADCG) method's ability to accurately localize fluorophores in diffraction-limited single molecule images (SMI) and analyze the impact of pre-processing and post-processing modules on ADCG's fluorophore localization. A synthetic dataset obtained from the 2013 Grand Challenge loc
This paper proposes and investigates a Bayesian relation between optimal L2 regularization strengths and the number of training patterns and hidden nodes used for an artificial neural network. The results support the proposed dependence for number of training patterns, while the dependence on hidden architecture was less clear. Finally, applying different regularization strengths on different laye
Fitting a model to a time-dependent ensemble average is a process repeated frequently throughout biophysics. A selected ensemble-averaged observable (⟨y(t)⟩) for a given system can be predicted through the use of an estimated ensemble average, where the estimated ensemble average is created via simulated or experimental data sets. Fitting a model to this estimated ensemble average allows for estim
Genetic algorithms are complex constructs often used as heuristic search methods in contexts ranging from combinatorial optimisation to in silico evolution. They draw inspiration from the principles of biological evolution by utilizing the concepts of mutation, reproduction and selection in order to improve a population of solutions. The solutions are often represented as abstract data sequences,
In this project, we compare two error functions for the purpose of training artificial neural networks on heavily censored data (data where key information is missing). J. Kalderstam et al. has shown that it is possible to train artificial neural networks directly on Harrell's C index \cite{Harrell} using genetic algorithms. He has also investigated the possibilities of improving the performa
During the last few years, crowding effects on the physics of proteins has become an increasingly popular topic of research. This is is because most biological processes involving proteins naturally take place in a crowded environment, e.g. in the cellular environment where macromolecules may occupy 30% of the volume. One such biological process would be the formation of amyloid aggregates, which
When using antibody microarrays to diagnose diseases, the process of quality control of the microarray data is an important step. Currently, a part of this process is performed manually by visual inspection. In this master project, we aim to automate the quality control in order to make it reproducible, as well as to find out which properties of the data that have the most to do with its quality.
Background: Worldwide, insulin-dependent type 1 diabetes is one of the most frequently diagnosed long-term endocrine disorders found in children and the incidences of this diseased is still increasing. In Sweden the routines are, according to national guidelines, when the child is diagnosed with type 1 diabetes, the child and its family remains at the hospital for about two weeks. There is limited
The purpose of this report is to evaluate the safety of life during evacuation from the icebreaker Sankt Erik in Stockholm in the event of fire onboard. Sankt Erik was decommissioned in 1977 and operates since 1980 as a museum under the Swedish Maritime Museum in Stockholm, moored at Galärvarvet. Being a museum no major alterations to the structure are allowed. Furthermore, the majority of the mat
Swedish legislation does not contain an explicit regulation regarding legal guardians’ liability to pay indemnity when children cause damage. The legal guardian has a supervision duty to observe, if violating this intentionally or by either negligence or willfulness the legal guardian risks liability for damages according to public liability. In 2010, the legislature chose to clarify the responsi
