Explore the vast range of titles available.

Purpose Some health providers ration knee arthroplasty on the basis of body mass index (BMI). There is no long-term data on the outcome of medial mobile-bearing unicompartmental knee arthroplasty (UKA) in different BMI groups. This study aimed to determine the effect of patient body mass index (BMI) on patient-reported outcomes and long-term survival of medial UKA in a large non-registry cohort. Our hypothesis is that increasing BMI would be associated with worse outcomes. Methods Data were analysed from a prospective cohort of 1000 consecutive medial mobile-bearing Oxford UKA with mean 10-year follow-up. Patients were grouped: BMI < 25, BMI 25 to < 30, BMI 30 to < 35 and BMI 35+. Oxford Knee Score (OKS) and Tegner Activity Score were assessed at 1, 5 and 10 years. Kaplan–Meier survivorship was calculated and compared between BMI groups. Results All groups had significant improvement in OKS and Tegner scores. BMI 35 + kg/m 2 experienced the greatest overall increase in mean OKS of 17.3 points ( p  = 0.02). There was no significant difference in ten-year survival, which was, from lowest BMI group to highest 92%, 95%, 94% and 93%. Conclusion There was no difference in implant survival between groups, and although there was no consistent trend in postoperative OKS, the BMI 35+ group benefited the most from UKA. Therefore, when UKA is used for appropriate indications, high BMI should not be considered to be a contraindication. Furthermore rationing based on BMI seems unjustified, particularly when the commonest threshold (BMI 35) is used. Level of evidence III.

Many real-world applications require artificial agents to compete and coordinate with other agents in complex environments. As a stepping stone to this goal, the domain of StarCraft has emerged as an important challenge for artificial intelligence research, owing to its iconic and enduring status among the most difficult professional esports and its relevance to the real world in terms of its raw complexity and multi-agent challenges. Over the course of a decade and numerous competitions 1 – 3 , the strongest agents have simplified important aspects of the game, utilized superhuman capabilities, or employed hand-crafted sub-systems 4 . Despite these advantages, no previous agent has come close to matching the overall skill of top StarCraft players. We chose to address the challenge of StarCraft using general-purpose learning methods that are in principle applicable to other complex domains: a multi-agent reinforcement learning algorithm that uses data from both human and agent games within a diverse league of continually adapting strategies and counter-strategies, each represented by deep neural networks 5 , 6 . We evaluated our agent, AlphaStar, in the full game of StarCraft II, through a series of online games against human players. AlphaStar was rated at Grandmaster level for all three StarCraft races and above 99.8% of officially ranked human players. AlphaStar uses a multi-agent reinforcement learning algorithm and has reached Grandmaster level, ranking among the top 0.2% of human players for the real-time strategy game StarCraft II.

Floodplain landscapes play a significant role in hydrologic fluxes, including connectivity to the alluvial aquifer and the biogeochemical processing of solutes from irrigation return flows. Variable spatial extents and limited temporal occurrence of active floodplains make quantifying their hydrologic and biogeochemical impacts problematic. To investigate, a surface-subsurface modeling practice was implemented to simulate hydrologic processes at the watershed scale in the heavily managed Lower Arkansas River Valley (LARV) (Colorado, USA). The SWAT+ model accounts for spatial variability of landscape features while simulating the fundamental physical principles that govern hydrologic processes within a watershed, such as runoff, infiltration, soil water routing, crop uptake, soil lateral flow, groundwater storage and flow, and streamflow. Using the gwflow module of SWAT+ simulates groundwater head, storage, and fluxes in response to hydrology and irrigation at the surface, replacing the original groundwater module. The primary objective of this thesis is to improve the implementation of floodplain landscapes in a modified version of SWAT+ with gwflow; and to assess the role of floodplains in aquifer recharge in the LARV. The model is run for the 1992-2020 period, with fifteen parameters calibrated for streamflow. Most years, flooding is insignificant in the managed LARV, and few floodplain-linked cells become active along the Arkansas River corridor. Flood scenarios for 100-year and 500-year events were run to observe the effects of including floodplain-exchange in SWAT+ gwflow models. Water balances reveal that the hydrologic process with the largest daily groundwater flux may occur through active floodplains, with implications on the annual change in storage for an aquifer system. Groundwater contributes 13% to streamflows through the standard simulation period. During the month of a 500-year flood scenario, groundwater produced 8% of surface flows, up from 3% without integrating floodplains. Activating floodplains in the 500-year flood scenario provided an additional flux, increasing groundwater storage by 3.5 % that year. Results are largely dependent on how floodplain landscapes are delineated; findings show that incorporating floodplains and floodplain-channel interaction into models likely brings the simulation into a stronger accordance with the real stream-aquifer system, as seasonal groundwater head and fluxes (groundwater saturation excess flow, groundwater evapotranspiration, groundwater return flows) are influenced by river water seeping to the aquifer during periods of flooding. Doing so allows the LARV model, and other models that use the floodplain option, to be used for quantifying the effects of flooding events on hydrological processes, nutrient processes, and management of wetlands and cropping systems within the floodplain of a river valley.

Cyber Threat Intelligence is the knowledge required to protect personal computers, corporations, and critical infrastructure from cyber threat actors. With the modern world's reliance on internet-connected devices, Cyber Threat Intelligence is a necessity. A prominent component of Cyber Threat Intelligence is malware analysis, research conducted to protect computer networks from malware attacks. Malware analysis has seen an application of Artificial Intelligence (AI) in recent years. To combat the increasingly versatile and mutable modern malware, Machine Learning (ML) is now a popular and effective complement to the existing signature-based techniques for malware triage and identification. However, ML is also a readily available tool for adversaries. Through adversarial learning on malware, adversaries have developed techniques for bypassing ML-based models by making their malware appear benign. Two challenges that have arisen in this area of study are modified malware detection and malware family classification. In this thesis, we aim to provide Deep Learning (DL) based solutions to these complex challenges. First, we propose H4rm0ny, the first Reinforcement Learning (RL) two-player game for malware generation and detection. Then, we propose Ch4os, a method for creating adversarial bytes with a generative framework. We also propose a practical and efficient solution for zero-day malware variant matching with reconstruction. Finally, we propose Mecha, a neuro-symbolic approach to open-set malware family classification. We have conducted multiple experiments to observe the efficacy of our solutions against datasets of thousands of software samples. All solutions we discuss in this thesis have improved against the state-of-the-art in empirical testing.

A reduction in the long-term capital gains tax rate provides investors with new strategies to minimize taxes and protect investment gains. One such opportunity exists when an investor decides to sell a profitable stock with a holding period of less than one-year, resulting in short-term ordinary taxes. The investor would find it more beneficial to sell the stock after one-year lapses, resulting in lower long-term capital gain taxes, although the longer holding period exposes the investor to the uncertainty of stock price movement. A strategy to extend the holding period without excess risk would be to use the protective put option strategy, sometimes referred to as \"investment insurance\". The strategy involves the purchase of a put option to protect against the possible decline in the stock price, to take advantage of the lower long-term capital gains tax rate, and to preserve the upside potential of the stock. Pursuant to IRS Publication 550, the IRS does not allow the use of a protective put to extend the holding period on the same security considered for sale. Since the IRS does not allow a direct protective put hedge, this study will explore an alternative strategy involving the purchase of a put on a highly correlated investment to extend the holding period to recognize lower capital gains tax rates. The paper presents example situations when an investor benefits from utilizing the correlated protective put option strategy.

Facilities Management supports the primary functions of teaching and research at higher education institutions, but it is often viewed as an area of cost containment rather a tool for achieving strategic growth. This view often results in an inefficient FM structure. The study reviews the efficacy of key aspects of facilities management structure including centralisation, decentralisation, service level management, strategic alignment and sourcing strategies.A multiple case study was undertaken at the facilities management divisions at two universities based in Gauteng Province, South Africa. The universities were both large, multi-campus facilities and were chosen based on their differing approach to facilities management. The case study method was suitable for this study as it provided in depth knowledge and observation within the institutions. Structured interviews were conducted with 8 senior managers at the institutions and data was collected using a mixed method approach. The Delphi technique was conducted with these senior managers in order to achieve consensus on the most significant factors influencing the future direction of facilities management.The findings provide empirical insights of how the facilities management is structured, and which worked more effectively at these higher education institutions. Valuable insights were found regarding centralising and decentralising, service level management, and alignment with strategic goals. The research also found advantages and disadvantages in the various sourcing strategies, and such information would be valuable to the practitioners for the purposes of benchmarking and improvement in the industry. The research fulfils an identified need to improve the efficacy of facilities management at higher education institutions.In order to improve validity, the two universities selected had differing approaches to Facilities Management. However, the findings may lack generalisability due to the limitations of the case study methodology.

Programming is a powerful and ubiquitous problem-solving tool. Developing systems that can assist programmers or even generate programs independently could make programming more productive and accessible, yet so far incorporating innovations in AI has proven challenging. Recent large-scale language models have demonstrated an impressive ability to generate code, and are now able to complete simple programming tasks. However, these models still perform poorly when evaluated on more complex, unseen problems that require problem-solving skills beyond simply translating instructions into code. For example, competitive programming problems which require an understanding of algorithms and complex natural language remain extremely challenging. To address this gap, we introduce AlphaCode, a system for code generation that can create novel solutions to these problems that require deeper reasoning. In simulated evaluations on recent programming competitions on the Codeforces platform, AlphaCode achieved on average a ranking of top 54.3% in competitions with more than 5,000 participants. We found that three key components were critical to achieve good and reliable performance: (1) an extensive and clean competitive programming dataset for training and evaluation, (2) large and efficient-to-sample transformer-based architectures, and (3) large-scale model sampling to explore the search space, followed by filtering based on program behavior to a small set of submissions.

Although methylenetetrahydrofolate reductase, a folate enzyme gene, has been associated with idiopathic male infertility, few studies have examined other folate-related metabolites and genes. We investigated whether idiopathic male infertility is associated with variants in folate, vitamin B12 （B12） and total homocysteine （tHcy）-related genes and measured these metabolites in blood. We conducted a case-control study that included 153 men with idiopathic infertility and 184 fertile male controls recruited at the Fertility Center and Antenatal Care Center, University Hospital, Malmo and Lund, Sweden. Serum folate, red cell folate （RCF）, serum B12, plasma tHcy and semen quality were measured. Subjects were genotyped for 20 common variants in 12 genes related to folate/B12/ homocysteine metabolism. Metabolite concentrations and genotype distributions were compared between cases and controls using linear and logistic regression with adjustment for covariates. The phosphatidylethanolamine N-methyltransferase （PEMT） M 175V and TCblR rs173665 polymorphisms were significantly associated with infertility （P=0.01 and P=0.009, respectively）, but not with semen quality. Among non-users of supplements, infertile men had lower serum folate concentrations than fertile men （12.89 vs. 14.73 nmoll^- 1 P=0.02）, but there were no significant differences in RCF, B 12 or tHcy. Folate, B 12 and tHcy concentrations were not correlated with any semen parameters. This study provides little support for low folate or B12 status in the pathogenesis of idiopathic male infertility. Although additional data are needed to confirm these initial findings, our results suggest that PEMTand TCbIR, genes involved in choline and B12 metabolism, merit further investigation in idiopathic male infertility.

Training of modern large neural networks (NN) requires a combination of parallelization strategies encompassing data, model, or optimizer sharding. When strategies increase in complexity, it becomes necessary for partitioning tools to be 1) expressive, allowing the composition of simpler strategies, and 2) predictable to estimate performance analytically. We present PartIR, our design for a NN partitioning system. PartIR is focused on an incremental approach to rewriting and is hardware-and-runtime agnostic. We present a simple but powerful API for composing sharding strategies and a simulator to validate them. The process is driven by high-level programmer-issued partitioning tactics, which can be both manual and automatic. Importantly, the tactics are specified separately from the model code, making them easy to change. We evaluate PartIR on several different models to demonstrate its predictability, expressibility, and ability to reach peak performance..