The promise of unsupervised learning methods lies in their potential to use vast amounts of unlabeled data to learn complex, highly nonlinear models with millions of free parameters. We consider two well-known unsupervised learning models, deep belief networks (DBNs) and sparse coding, that have recently been applied to a flurry of machine learning applications (Hinton & Salakhutdinov, 2006;Raina et al., 2007). Unfortunately, current learning algorithms for both models are too slow for large-scale applications, forcing researchers to focus on smaller-scale models, or to use fewer training examples.In this paper, we suggest massively parallel methods to help resolve these problems. We argue that modern graphics processors far surpass the computational capabilities of multicore CPUs, and have the potential to revolutionize the applicability of deep unsupervised learning methods. We develop general principles for massively parallelizing unsupervised learning tasks using graphics processors. We show that these principles can be applied to successfully scaling up learning algorithms for both DBNs and sparse coding. Our implementation of DBN learning is up to 70 times faster than a dual-core CPU implementation for large models. For example, we are able to reduce the time required to learn a four-layer DBN with 100 million free parameters from several weeks to around a single day. For sparse coding, we develop a simple, inherently parallel algorithm, that leads to a 5 to 15-fold speedup over previous methods.
Optic neuropathy has a broad differential diagnosis, including demyelinating disease, infection, and inflammatory etiologies such as sarcoidosis, radiation, and drug toxicity. Tacrolimus is a rare cause of optic neuropathy, but radiologists can benefit patients by suggesting the diagnosis.
We report a case of tacrolimus optic neuropathy initially suggested on the basis of MR imaging findings. The patient improved clinically and had near resolution of MR imaging findings after treatment, which has not been previously reported.
To understand the clinical practice pattern of general ophthalmologists in the management of retinal diseases. Also, aimed to explore the ophthalmologist’s perspective towards patient compliance and unmet need in the management of neovascular age-related macular degeneration (nAMD).A total of 108 ophthalmologists participated in this cross-sectional questionnaire-based survey. A paper-based questionnaire with a tool of twelve questions, with response options ranging on a five-point Likert scale of ‘strongly agree’ to ‘strongly disagree’ was provided to participants.Out of 108, 95.4% ophthalmologists confirmed that they were commonly consulted for nAMD amongst the different retinal disorders (RDs). The majority of respondents (87%) confirmed that 60% or fewer patients continue the treatment for a year. About 81.5% of ophthalmologists stated that fluid (Intra-retinal fluid, Sub-retinal fluid) on optical coherence tomography (OCT) was an extremely important parameter for disease activity. The survey revealed that injection frequency was the factor for non-compliance in majority of (>50%) patients. More than 64% of respondents opined that improved efficacy (70.4%), reduced treatment burden (64.8%), and longer acting agents/sustained delivery (64.8%) are the most critical unmet needs for nAMD patients.Based on the findings, it can be concluded that, in addition to functional outcomes i.e. visual acuity, ophthalmologists also considered retinal fluid and central retinal thickness as important parameters for treatment-related decisions. Ophthalmologists suggested that there is a need to develop longer-acting agents with improved efficacy which may help in reducing treatment burden in nAMD management.Longer acting anti-vascular endothelial growth factor (VGEF) agents with improved efficacy may help in reducing the treatment burden in nAMD management.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.