Search for: All records

Content caching is vital for enhancing web server efficiency and reducing network congestion, particularly in platforms predicting user actions. Despite many studies conducted toimprove cache replacement strategies, there remains space for improvement. This paper introduces STRCacheML, a Machine Learning (ML) assisted Content Caching Policy. STRCacheML leverages available attributes within a platform to make intelligent cache replacement decisions offline. We have tested various Machine Learning and Deep Learning algorithms to adapt the one with the highest accuracy; we have integrated that algorithm into our cache replacement policy. This selected ML algorithm was employed to estimate the likelihood of cache objects being requested again, an essential factor in cache eviction scenarios. The IMDb dataset, constituting numerous videos with corresponding attributes, was utilized to conduct our experiment. The experimental section highlights our model’s efficacy, presenting comparative results compared to the established approaches based on raw cache hits and cache hit rates. 

This is the AmeriFlux version of the carbon flux data for the site US-Skr Shark River Slough (Tower SRS-6) Everglades. Site Description - The Florida Everglades Shark River Slough Mangrove Forest site is located along the Shark River in the western region of Everglades National Park. Also referred to as site SRS6 of the Florida Coastal Everglades LTER program, freshwater in the mangrove riverine floods the forest floor under a meter of water twice per day. Transgressive discharge of freshwater from the Shark river follows annual rainfall distributions between the wet and dry seasons. Hurricane Wilma struck the site in October of 2005 causing significant damage. The tower was offline until the following October in order to continue temporally consistent measurements. In post-hurricane conditions, ecosystem respiration rates and solar irradiance transfer increased. 2007- 2008 measurements indicate that these factors led to an decline in both annual -NEE and daily NEE from pre-hurricane conditions in 2004-2005. 

We examine a novel setting in which two parties have partial knowledge of the elements that make up a Markov Decision Process (MDP) and must cooperate to compute and execute an optimal policy for the problem constructed from those elements. This situation arises when one party wants to give a robot some task, but does not wish to divulge those details to a second party-while the second party possesses sensitive data about the robot's dynamics (information needed for planning). Both parties want the robot to perform the task successfully, but neither is willing to disclose any more information than is absolutely necessary. We utilize techniques from secure multi-party computation, combining primitives and algorithms to construct protocols that can compute an optimal policy while ensuring that the policy remains opaque by being split across both parties. To execute a split policy, we also give a protocol that enables the robot to determine what actions to trigger, while the second party guards against attempts to probe for information inconsistent with the policy's prescribed execution. In order to improve scalability, we find that basis functions and constraint sampling methods are useful in forming effective approximate MDPs. We report simulation results examining performance and precision, and assess the scaling properties of our Python implementation. We also describe a hardware proof-of-feasibility implementation using inexpensive physical robots, which, being a small-scale instance, can be solved directly. 

Insects perform essential roles within ecosystems and can be vulnerable to climate change because of their small body size and limited capacity to regulate body temperature. Several groups of insects, such as bees and flies, are important pollinators of wild and cultivated plants. However, aspects of their thermal biology remain poorly studied, which limits predictions of their responses to climate change. We assessed the critical thermal maximum (CTMax) of bees and flies visiting flowers in urban and periurban areas in tropical and subtropical regions of the Americas. We also assessed the effect of the foraging time of the day on CTMax. Overall, we found that bees displayed higher CTMax than flies. Flies foraging in the morning and afternoon displayed similar CTMax while bees in the morning displayed a higher CTMax than in the afternoon. The results of this study suggest differences in the vulnerability to climate change between these two major groups of pollinators, with flies being more at risk. 

Abstract. Conventional and recently developed approaches for estimating turbulent scalar fluxes under stable atmospheric conditions are evaluated, with a focus on gases for which fast sensors are not readily available. First, the relaxed eddy accumulation (REA) classical approach and a recently proposed mixing length parameterization, labeled A22, are tested against eddy-covariance computations. Using high-frequency measurements collected from two contrasting sites (the frozen tundra near Utqiaġvik, Alaska, and a sparsely vegetated grassland in Wendell, Idaho, during winter), it is shown that the REA and A22 models outperform the conventional Monin–Obukhov similarity theory (MOST) utilized widely to infer fluxes from mean gradients. Second, scenarios where slow trace gas sensors are the only viable option in field measurements are investigated using digital filtering applied to fast-response sensors to simulate their slow-response counterparts. With a filtered scalar signal, the observed filtered eddy-covariance fluxes are referred to here as large-eddy-covariance (LEC) fluxes. A virtual eddy accumulation (VEA) approach, akin to the REA model but not requiring a mechanical apparatus to separate the gas flows, is also formulated and tested. A22 outperforms VEA and LEC in predicting the observed unfiltered (total) eddy-covariance (EC) fluxes; however, VEA can still capture the LEC fluxes well. This finding motivates the introduction of a sensor response time correction into the VEA formulation to offset the effect of sensor filtering on the underestimated net averaged fluxes. The only needed parameter for this correction is the mean velocity at the instrument height, a surrogate of the advective timescale. The VEA approach is very suitable and simple to use with gas sensors of intermediate speed (∼ 0.5 to 1 Hz) and with conventional open- or closed-path setups.