The problem of making predictions from theories that have landscapes of possible low energy parameters is reviewed. Conditions for such a theory to yield falsifiable predictions for doable experiments are given. It is shown that the hypothesis of cosmological natural selection satisfies these conditions, thus showing that it is possible to continue to do physics on a landscape without invoking ...

Word primes have been successfully used in the past to facilitate the processing of other words (semantic priming), but also to modify mental states such as emotion, cognition and motivation (conceptual priming). This work documented that the direction of motivational drive can be successfully changed, but left open whether its magnitude can be influenced as well. To find out, we asked subjects...

Document classification tasks were primarily tackled at word level. Recent research that works with character-level inputs shows several benefits over word-level approaches such as natural incorporation of morphemes and better handling of rare words. We propose a neural network architecture that utilizes both convolution and recurrent layers to efficiently encode character inputs. We validate t...

This paper explores gait recognition for various walking speeds. Normal-constant speed is one of the assumptions being made in many current gait recognition techniques. However, some techniques do not scale well when certain gait conditions such as walking speed varies. We demonstrate the characteristics of time-distance gait parameters, stride length and cadence, with respect to walking speed ...

The nuclear matrix elements M of the neutrinoless double beta decay (0νββ) of most nuclei with known 2νββ-decay rates are systematically evaluated using the Quasiparticle Random Phase Approximation (QRPA) and Renormalized QRPA (RQRPA). The experimental 2νββ-decay rate is used to adjust the most relevant parameter, the strength of the particle-particle interaction. With such procedure the M valu...

The experimental technique of fluorescence light observation is used in current and planned air shower experiments that aim at understanding the origin of ultrahigh energy cosmic rays. In the fluorescence technique, the geometry of the shower is reconstructed from the correlation between arrival time and incident angle of the signals detected by the telescope. The calculation of the expected li...

Background Although chronic obstructive pulmonary disease is a known cause of secondary polycythemia with elevated erythropoietic demands in response to hypoxemia, anemia is relatively common in patients with chronic obstructive pulmonary disease and is related to increased mortality. However, little is currently known about the relationship between various iron parameters and disease severity ...

Gaussian processes reinforcement learning provides an appealing framework for training the dialogue policy as it takes into account correlations of the objective function given different dialogue belief states, which can significantly speed up the learning. These correlations are modelled by the kernel function which may depend on hyper-parameters. So far, for real-world dialogue systems the hy...

This paper discusses and analyzes risk measure properties in order to understand how a risk measure has to be used to optimize the investor’s portfolio choices. In particular, we distinguish between two admissible classes of risk measures proposed in the portfolio literature: safety risk measures and dispersion measures. We study and describe how the risk could depend on other distributional pa...

Short Summary (in English) 66 Short Summary (in Hungarian) 67 ii Acknowledgements First of all, I would like to express my thanks to my supervisor András L®rincz for the continuous support and trust. I'm deeply indebted to the members of his research group for the friendly atmosphere and for their readiness to help. Namely, to Zoltán Bárdosi, István Szita and Bálint Takács. I'm especially grate...