One of the main reasons for protein clustering is prediction of structure, function and evolution. Many of current tools have disadvantage of high computational complexity due to all-to-all sequence alignment. If any tool works faster, it does not reach accuracy as other tools. Further disadvantage is processing on higher rate of similarity but homologous proteins can be similar with less identity. The process of clustering often ends when reach the condition which does not reflect sufficient quality of clusters. Master's thesis describes the design and implementation of new tool for clustering of protein sequences. New tool should not be computationally demanding but it should preserve required accuracy and produce better clusters. The thesis also describes testing of designed tool, evaluation of results and possibilities of its further development.

The Master’s thesis is focused on development of Android application which is able to recognize grid and clue text of swedish crossword and find a solution for it in crossword dictionary. The thesis proposes a process of tile segmentation using corner detection, recognizes text using Tesseract OCR and searches for solutions in local database. Developed application is tested on a gallery of photos captured using a mobile phone. The tile segmentation and solution searching provide good results, the rest of the process provides unsatisfactory results thanks to inaccurate OCR output.

No description available.

The problem of epipolar geometry estimation together with correspondence establishment in case of wide baseline and large scale changes and rotation has been addressed in this work. This work deals with cases that are heavily noised by outliers. The jump diffusion MCMC method has been employed to search for the non-degenerate epipolar geometry with the highest probabilistic support of putative correspondences. At the same time, inliers in the putative set are also identified. The jump steps involve large movements guided by a distribution of similarity based priors while diffusion steps are small movements guided by a distribution of likelihoods given by the Joint Feature Distribution (JFD). The 'best so far' samples are accepted in accordance to Metropolis-Hastings method. The diffusion steps are carried out by sampling conditioned on the 'best so far', making it local to the 'best so far' sample, while jump steps remain unconditioned and span across the correspondence and motion space according to a similarity based proposal distribution making large movements. We advance the theory in three novel ways. First, a similarity based prior proposal distribution which guide jump steps. Second, JFD based likelihoods which guide diffusion steps allowing more focused correspondence establishment while searching for epipolar geometry. Third, a measure of degeneracy that allows to rule out degenerate configurations. The jump diffusion framework thus defined allows handling over 90% outliers even in cases where the number of inliers is very few. Practically, the advancement lies in higher precision and accuracy that has been detailed in this work by comparisons. In this work, BLOGS is compared with LO-RANSAC, NAPSAC, MAPSAC and BEEM algorithm, which are the current state of the art competing methods, on a dataset that has significantly more change in baseline, rotation, and scale than those used in the state of the art. Performance of these algorithms and BLOGS are quantitatively benchmark for a comparison by estimating the error in the epipolar geometry given by root mean Sampson's distance from manually specified corresponding point pairs which serve as a ground truth. Not just is BLOGS able to tolerate very high outlier rates, but also gives result of similar quality in 10 times lesser number of iterations than the most competitive among the compared algorithms.

Process model matchers automate the detection of activities that represent similar functionality in different models. Thus, they provide support for various tasks related to the management of business processes including model collection management and process design. Yet, prior research primarily demonstrated the matchers’ effectiveness, i.e., the accuracy and the completeness of the results. In this context (i) the size of the empirical data is often small, (ii) all data is used for the matcher development, and (iii) the validity of the design decisions is not studied. As a result, existing matchers yield a varying and typically low effectiveness when applied to different datasets, as among others demonstrated by the process model matching contests in 2013 and 2015. With this in mind, the thesis studies the effectiveness of matchers by separating development from evaluation data and by empirically analyzing the validity and the limitations of design decisions. In particular, the thesis develops matchers that rely on different sources of information. First, the activity labels are considered as natural-language descriptions and the Bag-of-Words Technique is introduced which achieves a high effectiveness in comparison to the state of the art. Second, the Order Preserving Bag-of-Words Technique analyzes temporal dependencies between activities in order to automatically configure the Bag-of-Words Technique and to improve its effectiveness. Third, expert feedback is used to adapt the matchers to the domain characteristics of process model collections. Here, the Adaptive Bag-of-Words Technique is introduced which outperforms the state-of-the-art matchers and the other matchers from this thesis.

Word embeddings are useful in many tasks in Natural Language Processing and Information Retrieval, such as text mining and classification, sentiment analysis, sentence completion, or dictionary construction. Word2vec and its predecessor fastText, both well-known models to produce word embeddings, are powerful techniques to study the syntactic and semantic relations between words by representing them in a low-dimensional vector. By applying algebraic operations on these vectors semantic relationships such as word analogies, gender-inflections, or geographical relationships can be easily recovered. The aim of this work is to investigate how word embeddings could be utilized to augment and enrich queries in DBMSs, e.g. to compare text values according to their semantic relation or to group rows according to the similarity of their text values. For this purpose, we use pre-trained word embedding models of large text corpora such as Wikipedia. By exploiting this external knowledge during query processing we are able to apply inductive reasoning on text values. Thereby, we reduce the demand for explicit knowledge in database systems. In the context of the IMDB database schema, this allows for example to query movies that are semantically close to genres such as historical fiction or road movie without maintaining this information. Another example query is sketched in Listing 1, that returns the top-3 nearest neighbors (NN) of each movie in IMDB. Given the movie “Godfather” as input this results in “Scarface”, “Goodfellas” and “Untouchables”.

Description Logics (DLs) are a family of logic-based knowledge representation languages used to describe the knowledge of an application domain and reason about it in formally well-defined way. They allow users to describe the important notions and classes of the knowledge domain as concepts, which formalize the necessary and sufficient conditions for individual objects to belong to that concept. A variety of different DLs exist, differing in the set of properties one can use to express concepts, the so-called concept constructors, as well as the set of axioms available to describe the relations between concepts or individuals. However, all classical DLs have in common that they can only express exact knowledge, and correspondingly only allow exact inferences. Either we can infer that some individual belongs to a concept, or we can't, there is no in-between. In practice though, knowledge is rarely exact. Many definitions have their exceptions or are vaguely formulated in the first place, and people might not only be interested in exact answers, but also in alternatives that are "close enough". This thesis is aimed at tackling how to express that something "close enough", and how to integrate this notion into the formalism of Description Logics. To this end, we will use the notion of similarity and dissimilarity measures as a way to quantify how close exactly two concepts are. We will look at how useful measures can be defined in the context of DLs, and how they can be incorporated into the formal framework in order to generalize it. In particular, we will look closer at two applications of thus measures to DLs: Relaxed instance queries will incorporate a similarity measure in order to not just give the exact answer to some query, but all answers that are reasonably similar. Prototypical definitions on the other hand use a measure of dissimilarity or distance between concepts in order to allow the definitions of and reasoning with concepts that capture not just those individuals that satisfy exactly the stated properties, but also those that are "close enough".

In order to analyse body composition of MRI (Magnetic Resonance Imaging) volumes, atlas-based segmentation is often used to retrieve information from specific organs or anatomical regions. The method behind this technique is to use an already segmented image volume, an atlas, to segment a target image volume by registering the volumes to each other. During this registration a deformation field will be calculated, which is applied to a segmented part of the atlas, resulting in the same anatomical segmentation in the target. The drawback with this method is that the quality of the segmentation is highly dependent on the similarity between the target and the atlas, which means that many atlases are needed to obtain good segmentation results in large sets of MRI volumes. One potential solution to overcome this problem is to create the deformation field between a target and an atlas as a sequence of small deformations between more similar bodies.  In this master thesis a new method for atlas-based segmentation has been developed, with the anticipation of obtaining good segmentation results regardless of the level of similarity between the target and the atlas. In order to do so, 4000 MRI volumes were used to create a manifold of human bodies, which represented a large variety of different body types. These MRI volumes were compared to each other and the calculated similarities were saved in matrices called similarity models. Three different similarity measures were used to create the models which resulted in three different versions of the model. In order to test the hypothesis of achieving good segmentation results when the deformation field was constructed as a sequence of small deformations, the similarity models were used to find the shortest path (the path with the least dissimilarity) between a target and an atlas in the manifold.  In order to evaluate the constructed similarity models, three MRI volumes were chosen as atlases and 100 MRI volumes were randomly picked to be used as targets. The shortest paths between these volumes were used to create the deformation fields as a sequence of small deformations. The created fields were then used to segment the anatomical regions ASAT (abdominal subcutaneous adipose tissue), LPT (left posterior thigh) and VAT (visceral adipose tissue). The segmentation performance was measured with Dice Index, where segmentations constructed at AMRA Medical AB were used as ground truth. In order to put the results in relation to another segmentation method, direct deformation fields between the targets and the atlases were also created and the segmentation results were compared to the ground truth with the Dice Index. Two different types of transformation methods, one non-parametric and one affine transformation, were used to create the deformation fields in this master thesis. The evaluation showed that good segmentation results can be achieved for the segmentation of VAT for one of the constructed similarity models. These results were obtained when a non-parametric registration method was used to create the deformation fields. In order to achieve similar results for an affine registration and to improve the segmentation of other anatomical regions, further investigations are needed.

The control of type-I error is a focal point for clinical trials. On the other hand, it is also critical to be able to detect a truly efficacious treatment in a clinical trial. With recent success in supervised learning (classification and regression problems), artificial intelligence (AI) and machine learning (ML) can play a vital role in identifying efficacious new treatments. However, the high performance of the AI methods, particularly the deep learning neural networks, requires a much larger dataset than those we commonly see in clinical trials. It is desirable to develop a new ML method that performs well with a small sample size (ranges from 20 to 200) and has advantages as compared with the classic statistical models and some of the most relevant ML methods. In this dissertation, we propose a Similarity-Principle-Based Machine Learning (SBML) method based on the similarity principle assuming that identical or similar subjects should behave in a similar manner. SBML method introduces the attribute-scaling factors at the training stage so that the relative importance of different attributes can be objectively determined in the similarity measures. In addition, the gradient method is used in learning / training in order to update the attribute-scaling factors. The method is novel as far as we know. We first evaluate SBML for continuous outcomes, especially when the sample size is small, and investigate the effects of various tuning parameters on the performance of SBML. Simulations show that SBML achieves better predictions in terms of mean squared errors or misclassification error rates for various situations under consideration than conventional statistical methods, such as full linear models, optimal or ridge regressions and mixed effect models, as well as ML methods including kernel and decision tree methods. We also extend and show how SBML can be flexibly applied to binary outcomes. Through numerical and simulation studies, we confirm that SBML performs well compared to classical statistical methods, even when the sample size is small and in the presence of unmeasured predictors and/or noise variables. Although SBML performs well with small sample sizes, it may not be computationally efficient for large sample sizes. Therefore, we propose Recursive SBML (RSBML), which can save computing time, with some tradeoffs for accuracy. In this sense, RSBML can also be viewed as a combination of unsupervised learning (dimension reduction) and supervised learning (prediction). Recursive learning resembles the natural human way of learning. It is an efficient way of learning from complicated large data. Based on the simulation results, RSBML performs much faster than SBML with reasonable accuracy for large sample sizes.

Virtual screening is a well-established part of computer-aided drug design, which heavily employs similarity search and similarity modeling methods. Most of the popular methods are target agnostic, leaving space for design of new methods that would take into account the specifics of the particular molecular target. Additionally, newly developed methods suffer from two related issues: benchmarking and availability. Benchmarking in the domain often suffers from the use of inappropriate reference methods, lack of reproducibility, and the use of nonstandard benchmark datasets. Although there have been several benchmarking studies in the domain that aim at addressing these issues, mainly by offering a standardized comparison, they often suffer from similar drawbacks. For these reasons, new methods fail to gain trust and therefore fail to become a part of the standard toolbox, which thus consists mostly of older methods. In this work, we address the above-described issues. First, we introduce new adaptive methods for virtual screening. Then, to make our and other newly developed methods readily available, we have designed and implemented a virtual screening tool. To address the benchmarking issue, we have compiled a publicly available collection of benchmarking datasets and proposed a platform offering a...