· · · Repository hosted by TU Delft Library

We present the design of an algorithm for use in an interactivemusic system that automatically generates music playlists that fit the music preferences of a user. To this end, we introduce a formal model, define the problem of automatic playlist generation (APG), and proof its NP-hardness. We use a local search (LS)procedure employing a heuristic improvement to standard simulated annealing (SA) to solve the APG problem. In order to employ this LS procedure, we introduce an optimization variant of the APG problem, which includes the definition of penalty functions and a neighborhood structure. To improve upon the performance of the standard SA algorithm, we incorporated three heuristics referred to as song domain reduction, partial constraint voting, an da two-level neighborhood structure. We evaluate the developed algorithm by comparing it to a previously developed approach based on constraint satisfaction (CS),both in terms of run time performance and quality of the solutions. For the latter we not only considered the penalty of the resulting solutions, but we also performed a conclusive user evaluation to assess the subjective quality of the playlists generated by both algorithms. In all tests, the LS algorithm was shown to be a dramatic improvement over the CS algorithm.

Video processing in software is often characterized by highly fluctuating, content-dependent processing times, and a limited tolerance for deadline misses. We present an approach that allows close-to-average-case resource allocation to a single video processing task, based on asynchronous, scalable processing, and QoS adaptation. The QoS adaptation balances different QoS parameters that can be tuned by user-perception experiments: picture quality, deadline misses, and quality changes. We model the balancing problem as a discrete stochastic decision problem, and propose two closely related solution strategies, for which the processing-time statistics are determined off line and at run time, respectively. We enhance both strategies with a compensation for structural (non-stochastic) load fluctuations. Finally, we validate our approach by means of simulation experiments, and conclude that both enhanced strategies perform close to the theoretical optimum.

Motivation: In recent years increasing evidence appeared that breastcancer may not constitute a single disease at the molecular level,but comprises a heterogeneous set of subtypes. This suggests that instead of building a single predictor, better predictors might be constructed that solely target samples of a designated subtype. An unavoidable drawback of developing subtype-specific predictors, however,is that a stratification by subtype drastically reduces the numberof samples available for their construction. It is therefore questionable whether the potential benefit of subtyping can outweigh the drawback of a severe loss in sample size. Factors like unequal class distributions and differences in the number of samples per subtype, further complicate comparisons. Results: We present several evaluation strategies that facilitate a comprehensive comparison between subtype-specific predictors and predictors that do not take subtype information into account. Emphasis lies on careful control of sample size as well as class and subtype distributions. The methodology is applied to a large breast cancer compendium involving over 1500 arrays,using a state-of-the-art subtyping scheme. We show that the resulting subtype-specific predictors outperform those that do not take subtype information into account, especially when taking sample size considerations into account.

In this study we report on initial efforts to discover putative biomarkers for differential diagnosis of a systemic inflammatory response syndrome (SIRS) vs. sepsis; and different stages of sepsis. In addition, we also investigated whether there are proteins that can discriminate between patients who survived sepsis from those who did not. Our study group consisted of 16 patients, of which 6 died and 10 survived. We daily measured 28 plasma proteins, for the whole stay of the patients in the ICU. We observed that metalloproteinases and sE-selectin play a role in the distinction between SIRS and sepsis, and that IL-1a, IP-10 and sTNF-R2 and sFAS appear to be indicative for the progression from sepsis to septic shock. A combined measurement of MMP-3, -10, IL-1a, IP-10, sIL-2R, sFas, sTNF-R1, sRAGE, GM-CSF,IL-1ß, and Eotaxin allows for a good separation of patients that survived from those that died (mortality prediction with a sensitivity of 79% and specificity of 86%). Correlation analysis suggests a novel interaction between IL-1a and IP-10. The marker panel is ready tobe verified in a validation study with or without therapeutic intervention.

We investigate the potential to enhance breast cancer event predictors by exploiting subtype information. We do this with a two-stage approach that first determines a sample's subtype using a recent module-driven approach, and secondly constructs a subtype-specific predictor to predict a metastasis event within five years. Our methodology is validated on a large compendium of microarray breast cancer datasets,including 43 replicate array pairs for assessing subtyping stability. Note that stratifying by subtype strongly reduces the training set sizes available to construct the individual predictors, which may decrease performance. Besides sample size, other factors likeunequal class distributions and differences in the number of samplesper subtype, easily obscure a fair comparison between subtype-specific predictors constructed on different subtypes, but also between subtype specific and subtype a-specific predictors. Therefore, we constructed a completely balanced experimental design, in which none ofthe above factors play a role and show that subtype-specific eventpredictors clearly outperform predictors that do not take subtype information into account.

The use of a priori knowledge in aligning targeted sequencing data is investigated using computational experiments. With conventional aligners such as Bowtie, BWA or MAQ, alignment is performed against the whole genome. Using an alignment method in which the genomic position information from the target capture is incorporated, alignment can be done to just the target region. Investigating the effect of realistic target size, read length, read redundancy, the amount of off-target reads and sequencing error rate, improvements of up to a factor of 8 +/- 0.3 in alignment speed are found using an implementation of the Needleman-Wunsch algorithm which makes use of direct stringcomparison. This results in a total alignment time in target sequencing of around 1 min.