Birmingham City University Open Access Repository
Not a member yet
8823 research outputs found
Sort by
Managerial Overconfidence and Corporate Investment Decision-Making: Theoretical and Empirical Studies
No full textChapter I of this PhD thesis is a literature review that systematically analyses the existing literature on behavioural and traditional finance. It critically examines traditional finance theories' limitations in explaining market irregularities and emphasizes the rising importance of behavioural finance. It highlights how conventional theories fall short in understanding market dynamics, paving the way for behavioural finance. The chapter explores overconfidence bias, a significant cognitive bias affecting financial decision-making, and debates whether it should be analysed separately as three distinct concepts (overprecision, overplacement and overestimation) or as a unified concept of overconfidence. It concludes that while behavioural finance complements traditional finance, it cannot replace it. Integrating insights from both fields is crucial for informed investment decisions. Regarding overconfidence bias, it notes challenges in accurately measuring its various aspects but advocates for studying them under a unified concept.
Chapter II introduces a theoretical model that utilizes algebraic expressions to analyse decision-making scenarios and assess the optimality of entrepreneurs' choices under different psychological states, including rationality, overconfidence, and pessimism. This chapter pioneers the examination of investment and organizational form selection through a behavioural finance lens. It introduces a model where entrepreneurs face three choices: launching an investment project with limited liability, initiating a project with unlimited liability, or refraining from action altogether. Entrepreneurs' decisions hinge on factors like projected project performance, their level of bias towards project expectations, and tax considerations encompassing both corporate and personal income taxes. When entrepreneurs exhibit overconfidence or pessimism, scenarios of overinvestment or underinvestment arise. The model suggests that effective government policy should account for the prevalent entrepreneurial bias. If confidence diminishes to breed pessimism, tax rates (both corporate and personal) should decrease; conversely, if overconfidence prevails, tax rates should rise. The study demonstrates that a combination of corporate tax tailored for limited liability businesses and a universal personal income tax could enhance decision-making efficiency and bolster social surplus.
The aim of Chapter III of the thesis is to analyse how CEO characteristics, managerial overconfidence and financial factors influence investment levels in UK SMEs. The dataset in use encompasses 256 UK SMEs, spanning the years 2014 to 2019. To measure CEO overconfidence, a proxy was created based on 2421 manually gathered articles depicting CEOs. The data has been analysed using the R programming language, employing four econometric models: Generalized Method of Moments (GMM), Fixed Effects, Random Effects, and Ordinary Least Squares (OLS) regressions. The study's findings emphasize the weight of CEO behavioural biases, especially overconfidence, and their personal attributes in shaping investment choices. Notably, a positive correlation emerges between CEO overconfidence and investment rates, signifying that highly confident CEOs tend to embrace more extensive investment. Similarly, CEO’s level of experience positively correlates with higher investment rates. In contrast, CEO tenure, age, and the count of directors exhibit negative relationships with investment levels. Of the economic factors considered, inflation was the sole variable that exhibited a noteworthy impact on the investment level
Neither an Individualised Nor a Standardised Sodium Bicarbonate Strategy Improved Performance in High-Intensity Repeated Swimming, or a Subsequent 200 m Swimming Time Trial in Highly Trained Female Swimmers
No full textInconsistent swimming performances are often observed following sodium bicarbonate (NaHCO3) ingestion, possibly because the time taken to reach peak blood buffering capacity is highly variable between individuals. Personalising NaHCO3 ingestion based on time-to-peak blood bicarbonate (HCO3−) could be a solution; however, this strategy is yet to be explored in swimming, or adequately compared to standardised NaHCO3 approaches. Therefore, six highly trained female swimmers ingested 0.3 g·kg BM−1 NaHCO3 in capsules to pre-determine their individual time-to-peak blood HCO3−. They then participated in three experimental trials, consisting of a 6 × 75 m repeated sprint swimming test, followed by a 200 m maximal time trial effort after 30 min active recovery. These experiments were conducted consuming a supplement at three different timings: individualised NaHCO3 (IND: 105–195 min pre-exercise); standardised NaHCO3 (STND: 150 min pre-exercise); and placebo (PLA: 90 min pre-exercise). Both NaHCO3 strategies produced similar increases in blood HCO3− prior to exercise (IND: +6.8 vs. STND: +6.1 mmol·L−1, p < 0.05 vs. PLA) and fully recovered blood HCO3− during active recovery (IND: +6.0 vs. STND: +6.3 mmol·L−1 vs. PLA, p < 0.05). However, there were no improvements in the mean 75 m swimming time (IND: 48.2 ± 4.8 vs. STND: 48.9 ± 5.8 vs. PLA: 49.1 ± 5.1 s, p = 0.302) nor 200 m maximal swimming (IND: 133.6 ± 5.0 vs. STND: 133.6 ± 4.7 vs. PLA: 133.3 ± 4.4 s, p = 0.746). Regardless of the ingestion strategy, NaHCO3 does not appear to improve exercise performance in highly trained female swimmers
Strategies for Resilience and Battery Life Extension in the Face of Communication Losses for Isolated Microgrids
No full textThis study addresses the challenges of energy deficiencies and high impact low probability (HILP) events in modern electrical grids by developing resilient microgrid energy management strategies. It introduces a sliding Model Predictive Control (MPC) methodology integrated with Battery Energy Storage Systems (BESS), emphasizing extending battery life and prioritizing critical loads during HILP events. This approach focuses on extending the sustainability of battery operation by linearizing the battery lifecycle within the optimization framework. Furthermore, this research proposed a straightforward method to mitigate communication disruptions during HILP events, thereby ensuring operational integrity. This focused approach enhances isolated microgrid resilience and sustainability, offering a strategic response to contemporary environmental challenges
Anomaly-Based Intrusion Detection Model Using Deep Learning for IoT Networks
No full textThe rapid growth of Internet of Things (IoT) devices has brought numerous benefits to the interconnected world. However, the ubiquitous nature of IoT networks exposes them to various security threats, including anomaly intrusion attacks. In addition, IoT devices generate a high volume of unstructured data. Traditional intrusion detection systems often struggle to cope with the unique characteristics of IoT networks, such as resource constraints and heterogeneous data sources. Given the unpredictable nature of network technologies and diverse intrusion methods, conventional machine-learning approaches seem to lack efficiency. Across numerous research domains, deep learning techniques have demonstrated their capability to precisely detect anomalies. This study designs and enhances a novel anomaly-based intrusion detection system (AIDS) for IoT networks. Firstly, a Sparse Autoencoder (SAE) is applied to reduce the high dimension and get a significant data representation by calculating the reconstructed error. Secondly, the Convolutional Neural Network (CNN) technique is employed to create a binary classification approach. The proposed SAE-CNN approach is validated using the Bot-IoT dataset. The proposed models exceed the performance of the existing deep learning approach in the literature with an accuracy of 99.9%, precision of 99.9%, recall of 100%, F1 of 99.9%, False Positive Rate (FPR) of 0.0003, and True Positive Rate (TPR) of 0.9992. In addition, alternative metrics, such as training and testing durations, indicated that SAE-CNN performs better
Artificial Intelligence Life Cycle: The Detection and Mitigation of Bias
No full textThe rapid expansion of Artificial Intelligence(AI) has outpaced the development of ethical guidelines and regulations, raising concerns about the potential for bias in AI systems. These biases in AI can manifest in real-world applications leading to unfair or discriminatory outcomes in areas like job hiring, loan approvals or criminal justice predictions. For example, a biased AI model used for loan prediction may deny loans to qualified applicants based on demographic factors such as race or gender. This paper investigates the presence and mitigation of bias in Machine Learning(ML) models trained on the Adult Census Income dataset, known to have limitations in gender and race. Through comprehensive data analysis, focusing on sensitive attributes like gender, race and relationship status, this research sheds light on complex relationships between societal biases and algorithmic outcomes and how societal biases can be rooted and amplified by ML algorithms. Utilising fairness metrics like demographic parity(DP) and equalised odds(EO), this paper quantifies the impact of bias on model predictions. The results demonstrated that biased datasets often lead to biased models even after applying pre-processing techniques. The effectiveness of mitigation techniques such as reweighting(Exponential Gradient(EG)) to reduce disparities was examined, resulting in a measurable reduction in bias disparities. However, these improvements came with trade-offs in accuracy and sometimes in other fairness metrics, identified the complex nature of bias mitigation and the need for precise consideration of ethical implications. The findings of this research highlight the critical importance of addressing bias at all stages of the AI life cycle, from data collection to model deployment. The limitation of this research, especially the use of EG, demonstrates the need for further development of bias mitigation techniques that can address complex relationships while maintaining accuracy. This paper concludes with recommendations for best practices in Artificial Intelligence development, emphasising the need for ongoing research and collaboration to mitigate bias by prioritising ethical considerations, transparency, explainability, and accountability to ensure fairness in AI systems
Talent inclusion: An imperfect solution to genetic testing in sport - Response to commentaries
No full textWe are extremely grateful our esteemed colleagues Craig Pickering, Duarte Araújo, Keith Davids, and Kevin Till have read and offered insightful reflections on the target article “Talent inclusion and genetic testing in sport: A practitioner’s guide”. We take the opportunity in the present article to respond to the three commentaries provided by these authors. In our target article, we highlighted at this moment in time, there is unequivocal disapproval in the scientific community with regards to the implementation of genetic testing in sport. Despite an insufficient evidence base, however, various stakeholders (e.g., athletes, support staff) have used, and will likely continue using, genetic tests. We offered potential explanations regarding the allure of genetic information to sports stakeholders before suggesting some imperfect solutions in terms of increasing genetic literacy, promoting talent inclusion, and following a minimum set of best practice guidelines
Hands-free Multi-Type Character Text Entry in Virtual Reality
No full textMulti-type characters, including uppercase and lowercase letters, symbols, and numbers, are essential in text entry activities. Although multi-type characters are used in passwords, instant messages, and document composition, there has been limited exploration of multi-character text entry for virtual reality head-mounted displays (VR HMDs). Typically, multi-type character entry requires four kinds of keyboards between which users need to switch. In this research, we explore hands-free approaches for rapid multi-type character entry. Our work explores two efficient and usable hands-free approaches for character selection: eye blinks and dwell. To enable quick switching between keyboards, we leverage the usability and efficiency of continuous head motions in the form of cross-based activation. In a pilot study, we explored the usability and efficiency of four locations of the switch keys, the two hands-free selection mechanisms, and crossing-based switching. In the main experiment, we evaluated four user-inspired layouts designed according to the findings from the pilot study. Results show that both blinking and dwell can work well with crossing-based switching and could lead to a relatively fast text entry rate (5.64 words-per-minute (WPM) with blinking and 5.42 WPM with dwell) with low errors (lower than 3% not corrected error rate (NCER)) for complex 8-digit passwords with upper/lowercase letters, symbols, and numbers. For sentences derived from the Brown Corpus, participants can reach 8.48 WPM with blinking and 7.78 WPM with dwell. Overall, as a first exploration, our results show that it is usable and efficient to perform hands-free text entry in VR using either eye blinks or dwell for character selection and crossing for mode switching
War crimes and questions on justice in asymmetric warfare: the case of Iraq
No full textThe concepts of human rights and international justice are products of the 20th century and of the so-called enlightened Western civilisation. As the 21st century began, the War on Terror was declared by the US–UK coalition, leading to hundreds of thousands of civilian deaths in the Middle East, as well as long-lasting insecurity in all areas of human life. Yet more than 20 years on, justice and accountability are still pending, when it comes to state violence, while a policy of “kill-not-capture” has dominated responses to violent non-state actors, such as Al-Qaeda and ISIS fighters, hampering any efforts at reconciliation and peace, and adding to regional and global injustice and insecurity
Reimagining Past Histories and Experiences through Performative Photography and Auto-ethnography
No full textThis article discusses an autoethnographic exploration within artistic research underpinned by performativity and temporality in photography. Inspired by a childhood photograph depicting the author's Danish grandfather, and a story about his ongoing chess through postal mail upon migrating to Argentina, the study explores the affective and performative power of photography and chess to re-imagine a personal narrative. By decentering the artist researcher’s voice, the work reveals the materiality and layers of temporal gap in the act of reimagining the past. This article contributes to the artistic through an innovative autoethnographic inquiry emerging within the transformative space integral to art practice
Random transformations to improve mitigation of query-based black-box attacks
No full textThis paper proposes methods to upstage the best-known defences against query-based black-box attacks. These benchmark defences incorporate gaussian noise into input data during inference to achieve state-of-the-art performance in protecting image classification models against the most advanced query-based black-box attacks. Even so there is a need to improve upon them; for example, the widely benchmarked Random noise defense (RND) method has demonstrated limited robustness – achieving only 53.5% and 18.1% with a ResNet-50 model on the CIFAR-10 and ImageNet datasets, respectively – against the square attack, which is commonly regarded as the state-of-the-art black-box attack. Therefore, in this work, we propose two alternatives to gaussian noise addition at inference time: random crop-resize and random rotation of the input images. Although these transformations are generally used for data augmentation while training to improve model invariance and generalisation, their protective potential against query-based black-box attacks at inference time is unexplored. Therefore, for the first time, we report that for such well-trained models either of the two transformations can also blunt powerful query-based black-box attacks when used at inference time on three popular datasets. The results show that the proposed randomised transformations outperform RND in terms of robust accuracy against a strong adversary that uses a high budget of 100,000 queries based on expectation over transformation (EOT) of 10, by 0.9% on the CIFAR-10 dataset, 9.4% on the ImageNet dataset and 1.6% on the Tiny ImageNet dataset. Crucially, in two even tougher attack settings, that is, high-confidence adversarial examples and EOT-50 adversary, these transformations are even more effective as the margin of improvement over the benchmarks increases further