Deep reinforcement learning is quickly changing the field of artificial intelligence. These models are able to capture a high level understanding of their environment, enabling them to learn difficult dynamic tasks in a variety of domains. In the database field, query optimization remains a difficult problem. Our goal in this work is to explore the capabilities of deep reinforcement learning in...

We show that two-band superconductors harbor hidden criticality deep in the superconducting state, stemming from the critical temperature of the weaker band taken as an independent system. For sufficiently small interband coupling γ the coherence length of the weaker band exhibits a remarkable deviation from the conventional monotonic increase with temperature, namely, a pronounced peak close t...

Twenty-one cases of a vegetative state (VS) caused by various kinds of brain damage were evaluated neurologically and electrophysiologically three months after brain injury. These cases were treated by deep brain stimulation (DBS) therapy, and followed up for over 10 years. The mesencephalic reticular formation was selected as a target in two cases, and the thalamic centre median-parafascicular...

Context-dependent phone models are used in modern speech recognition systems to account for co-articulation effects. Due to the vast number of possible context-dependent phones, statetying is typically used to reduce the number of target classes for acoustic modeling. We propose a novel approach for state-tying which is completely data dependent and requires no domain knowledge. Our method firs...

Such high energy processes as deep inelastic scattering (DIS) and Drell-Yan heavy lepton pair production are among the most important tools to probe the quark and gluon interaction. We calculated non-perturbative corrections to the LO cross section formulae for DIS and the Drell-Yan process. The interaction of partons was taken into account via dressing the incoming quark lines with spectral fu...

—In this work we propose a new deep learning tool – deep dictionary learning. Multi-level dictionaries are learnt in a greedy fashion – one layer at a time. This requires solving a simple (shallow) dictionary learning problem; the solution to this is well known. We apply the proposed technique on some benchmark deep learning datasets. We compare our results with other deep learning tools like s...

Surgical therapy for Parkinson's disease has a long history beginning in the 1930s with empirical exploration of different brain targets, such as resection of the primary motor cortex or extirpation of the caudate. Recently, there has been a renaissance of functional neurosurgery for the treatment of advanced Parkinson's disease, particularly deep brain stimulation (DBS). To date, DBS of the gl...

In this paper, we introduce a novel approach for diagnosis of Parkinson’s Disease (PD) based on deep Echo State Networks (ESNs). The identification of PD is performed by analyzing the whole time-series collected from a tablet device during the sketching of spiral tests, without the need for feature extraction and data preprocessing. We evaluated the proposed approach on a public dataset of spir...

At present, deep ultraviolet (DUV) lasers at the wavelength of fourth harmonics of 1 μm (266 nm/258 nm) and at the wavelength of 193 nm are widely utilized in science and industry. We review the generation of these DUV lasers by nonlinear frequency conversion processes using solid-state/fiber lasers as the fundamental frequency. A DUV laser at 258 nm by fourth harmonics generation (FHG) could a...

General arguments suggest that the non-perturbative background field in QCD may have a non-trivial spin structure. We discuss how this effect may be manifest in semi-inclusive measurements of fast pions in polarised deep inelastic scattering.