Human Bias Research Articles

Loan Approval Prediction System Using Machina Learning

With the growth of the banking sector, the identification of reliable borrowers that must maintain the natural core income and asset accumulation becomes a key issue. Despite all security measures, reliability of customers remains an unclear question. To tackle this barrier, banking management which is directed towards customer loan repayment consistency is required. Credit approval is a significant aspect of economy since it determines the allocation of credit-linked funds. Today, machine learning is known for its power to automate and scale up the processing of application for loans. This project will begin with data collection which will consist of data on historical (regarding the past) loan applications and the borrower profiles. The dataset has features of the credit score, income, previous work experience, debt-to-income ratio, and loan repayment record. This way, the models learn through the strengths to find good features and the reasons for accepting a loan. They are the experts in these areas and can forecast the potential patterns and connections of the data. Within the scope of this work, the supervised algorithms used are logistic regression, decision trees, random forests, and support vector machines. These algorithms are applied to the dataset available to often produce results like binary classification and regression. The adoption of machine learning among financial institutions is intended for a faster processing of loans which is their benefit. What is credit scoring, it is a tool which automate manual loan application review thereby increases efficiency. The machine-learning algorithms that analyse applications for loans could cut down on the possibilities of human biases and mistakes which are an inherent part of the process. Also, ML uses the model to recognize borrowers who may default and subsequently lower the likelihood of default. Part of the task involve utilizing historical credit market data and implementing ML algorithms to develop a highly accurate and reliable loan approving system based on trained-data, random forests, the stream of loans and reliable clients.

DeepLabCut-based daily behavioural and posture analysis in a cricket.

Circadian rhythms are indispensable intrinsic programs that regulate the daily rhythmicity of physiological processes, such as feeding and sleep. The cricket has been employed as a model organism for understanding the neural mechanisms underlying circadian rhythms in insects. However, previous studies measuring rhythm-controlled behaviours only analysed locomotive activity using seesaw-type and infrared sensor-based actometers. Meanwhile, advances in deep learning techniques have made it possible to analyse animal behaviour and posture using software that is devoid of human bias and does not require physical tagging of individual animals. Here, we present a system that can simultaneously quantify multiple behaviours in individual crickets - such as locomotor activity, feeding, and sleep-like states - in the long-term, using DeepLabCut, a supervised machine learning-based software for body keypoints labelling. Our system successfully labelled the six body parts of a single cricket with a high level of confidence and produced reliable data showing the diurnal rhythms of multiple behaviours. Our system also enabled the estimation of sleep-like states by focusing on posture, instead of immobility time, which is a conventional parameter. We anticipate that this system will provide an opportunity for simultaneous and automatic prediction of cricket behaviour and posture, facilitating the study of circadian rhythms.

Decoding violence against women: analysing harassment in middle eastern literature with machine learning and sentiment analysis

The rising prevalence of harassment in Middle Eastern countries is mirrored in literary works from the region. However, extracting data from these texts to understand the typology and frequency of the cases poses a significant challenge due to human cognitive limitations and potential biases. Thus, this study aims to use natural language processing (NLP) approaches to propose a machine learning framework for text mining of sexual harassment content in literary texts. The data source for this study consists of twelve Middle Eastern novels. The proposed framework involves the classification of physical and non-physical types of sexual harassment using a machine-learning model. Lexicon-based sentiment and emotion detection are applied to sentences containing instances of sexual harassment for data labelling and analysis. Finally, a long short-term memory-gated recurrent unit (LSTM-GRU) deep learning model is built to classify the sentiment characteristics that induce sexual harassment. The proposed model achieved an accuracy of 75.8% while outperforming five other algorithms. Additionally, a sentiment classification with three labels—negative, positive, and neutral—was developed using an LSTM-GRU RNN deep learning model. The accuracy of this model was 84.5%. Most statements, even those involving physical sexual harassment, which had greater levels of sexual harassment, had negative sentiments, according to lexicon-based sentiment analysis. This study contributes to the field of text mining by providing a novel approach to identifying instances of sexual harassment in literature in English from the Middle East. The use of machine learning models and sentiment analysis techniques allows for more accurate identification and classification of different types of sexual harassment. Furthermore, this study sheds light on the prevalence of sexual harassment in Middle Eastern countries and highlights the need for further research and action to address this issue.

Examining Ethical Aspects of AI: Addressing Bias and Equity in the Discipline

he rapid progress in implementing Artificial Intelligence (AI) across various domains such as healthcare decision-making, medical diagnosis, and others has raised significant concerns regarding the fairness and bias embedded within AI systems. This is particularly crucial in sectors like healthcare, employment, criminal justice, credit scoring, and the emerging field of generative AI models (GenAI) producing synthetic media. Such systems can lead to unfair outcomes and perpetuate existing inequalities, including biases ingrained in the synthetic data representation of individuals.This survey paper provides a concise yet comprehensive examination of fairness and bias in AI, encompassing their origins, ramifications, and potential mitigation strategies. We scrutinize sources of bias, including data, algorithmic, and human decision biases, shedding light on the emergent issue of generative AI bias where models may replicate and amplify societal stereotypes. Assessing the societal impact of biased AI systems, we spotlight the perpetuation of inequalities and the reinforcement of harmful stereotypes, especially as generative AI gains traction in shaping public perception through generated content.Various proposed mitigation strategies are explored, with an emphasis on the ethical considerations surrounding their implementation. We stress the necessity of interdisciplinary collaboration to ensure the effectiveness of these strategies. Through a systematic literature review spanning multiple academic disciplines, we define AI bias and its various types, delving into the nuances of generative AI bias. We discuss the adverse effects of AI bias on individuals and society, providing an overview of current approaches to mitigate bias, including data preprocessing, model selection, and post-processing. Unique challenges posed by generative AI models are highlighted, underscoring the importance of tailored strategies to address them effectively.Addressing bias in AI necessitates a holistic approach, involving diverse and representative datasets, enhanced transparency, and accountability in AI systems, and exploration of alternative AI paradigms prioritizing fairness and ethical considerations. This survey contributes to the ongoing discourse on developing fair and unbiased AI systems by outlining the sources, impacts, and mitigation strategies related to AI bias, with a particular focus on the burgeoning field of generative AI.

Examining trade routes through the Thai–Malay Peninsula: A simulation analysis

An important area for understanding human movement and trade routes during the Early Historic Period in Southeast Asia is the Thai–Malay Peninsula where external records date back 2300 years. Early east–west trade could be routed around the peninsula or sailing time could be reduced by taking terrestrial shortcuts across the peninsula. However, spatial research, particularly on transpeninsula routes, is insufficient to supplement the gaps between written historical records and excavated archaeological sites. This study aimed to simulate transpeninsula routes across the entire Thai–Malay Peninsula using a digital elevation model (DEM) to eliminate human biases when exploring the actual terrain. The simulation results reveal some intriguing characteristics of potential routes that can be used to divide the Thai–Malay Peninsula into five zones. This zonation is associated with external historical records and archaeological evidence before the twelfth century AD to assess the efficacy of different transpeninsula routes. These data are also utilized to propose potential areas of undiscovered archaeological sites within the Thai–Malay Peninsula.

Analysis of Arch Bridge Condition Data to Identify Network-Wide Controls and Trends

This paper outlines an initial analysis of 20 years of data held on an electronic bridge management database for approximately 3500 arch bridges across Northern Ireland (NI) by the Department for Infrastructure. Arch bridges represent the largest group of bridge types, making up nearly 56% of the total bridge stock in NI. This initial analysis aims to identify trends that might help inform maintenance decisions in the future. Consideration of the Bridge Condition Indicator (BCI) average value for the overall arch bridge stock indicates the potential for regional variations in the overall condition and the potential for human bias in inspections. The paper presents the most prevalent structural elements and associated defects recorded in the inspections of arch bridges. This indicated a link to scour and undermining for the worst-conditioned arch bridges. An Analysis of Variance (ANOVA) analysis identified function, number of spans, and deck width as significant factors during the various deterioration stages in a bridge’s lifecycle.

Few-shot Tumor Bud Segmentation Using Generative Model in Colorectal Carcinoma.

Current deep learning methods in histopathology are limited by the small amount of available data and time consumption in labeling the data. Colorectal cancer (CRC) tumor budding quantification performed using H&E-stained slides is crucial for cancer staging and prognosis but is subject to labor-intensive annotation and human bias. Thus, acquiring a large-scale, fully annotated dataset for training a tumor budding (TB) segmentation/detection system is difficult. Here, we present a DatasetGAN-based approach that can generate essentially an unlimited number of images with TB masks from a moderate number of unlabeled images and a few annotated images. The images generated by our model closely resemble the real colon tissue on H&E-stained slides. We test the performance of this model by training a downstream segmentation model, UNet++, on the generated images and masks. Our results show that the trained UNet++ model can achieve reasonable TB segmentation performance, especially at the instance level. This study demonstrates the potential of developing an annotation-efficient segmentation model for automatic TB detection and quantification.

Human bias in evaluating AI product creativity

This study investigated the evaluation behavior of products generated by artificial intelligence (AI) by manipulating the creativity of rating stimuli according to their high and low novelty and usefulness. In addition, we analyzed the effect of perception of the generative AI on creativity ratings. We found a bias toward lower creativity ratings only for highly useful products when they were identified as AI-generated, and this bias was particularly strong for individuals with greater perceived threat from the generative AI. Our research highlights the importance of product attribute quality and producer identity in creativity ratings, and provides implications for the integration of AI into creative industries.

Unaware and unaccepting: Human biases and the advent of artificial intelligence.

Unaware and unaccepting: Human biases and the advent of artificial intelligence.

Are the dynamics of wall turbulence in minimal channels and larger domain channels equivalent? A graph-theoretic approach

This work proposes two algorithmic approaches to extract critical dynamical mechanisms in wall-bounded turbulence with minimum human bias. In both approaches, multiple types of coherent structures are spatiotemporally tracked, resulting in a complex multilayer network. Network motif analysis, i.e., extracting dominant non-random elemental patterns within these networks, is used to identify the most dominant dynamical mechanisms. Both approaches, combined with network motif analysis, are used to answer whether the main dynamical mechanisms of a minimal flow unit (MFU) and a larger unconstrained channel flow, labeled a full channel (FC), at Reτ ≈ 180, are equivalent. The first approach tracks traditional coherent structures defined as low- and high-speed streaks, ejections, and sweeps. It is found that the roll-streak pairing, consistent with the current understanding of self-sustaining processes, is the most significant and simplest dynamical mechanism in both flows. However, the MFU has a timescale for this mechanism that is approximately 2.83 times slower than that of the FC. In the second approach, we use semi-Lagrangian wavepackets and define coherent structures from their energetic streak, roll, and small-scale phase space. This method also shows similar motifs for both the MFU and FC. It indicates that, on average, the most dominant phase-space motifs are similar between the two flows, with the significant events taking place approximately 2.21 times slower in the MFU than in the FC. This value is more consistent with the implied timescale ratio of only the slow speed streaks taking part in the roll-streak pairing extracted using the first multi-type spatiotemporal approach, which is approximately 2.17 slower in the MFU than the FC.

Uses and limitations of artificial intelligence for oncology.

Modern artificial intelligence (AI) tools built on high-dimensional patient data are reshaping oncology care, helping to improve goal-concordant care, decrease cancer mortality rates, and increase workflow efficiency and scope of care. However, data-related concerns and human biases that seep into algorithms during development and post-deployment phases affect performance in real-world settings, limiting the utility and safety of AI technology in oncology clinics. To this end, the authors review the current potential and limitations of predictive AI for cancer diagnosis and prognostication as well as of generative AI, specifically modern chatbots, which interfaces with patients and clinicians. They conclude the review with a discussion on ongoing challenges and regulatory opportunities in the field.

Revolutionizing Disease Diagnosis: Integrating Artificial Intelligence and IOT In Smart Healthcare Systems

This research paper delves into the transformative role of Artificial Intelligence (AI) and the Internet of Things (IoT) in revolutionizing disease diagnosis within smart healthcare systems. AI, a cornerstone in computer science, is now pivotal in healthcare, offering sophisticated algorithms for analyzing medical data, thus aiding in decision-making and predictions. IoT extends this capability by enabling continuous data collection through web-enabled devices, including wearables and implanted sensors. The integration of AI and IoT in smart healthcare systems not only refines medical procedures but also enhances patient experiences and operational workflows. This paper explores how AI-driven routines in conjunction with IoT data streams facilitate rapid and accurate disease diagnosis, surpassing traditional methods often limited by human biases. The adoption of these technologies in healthcare promises to improve diagnostic accuracy, expedite processes, and broaden access to medical expertise, especially in remote areas. However, challenges such as ensuring data privacy, security, and ethical use of AI remain paramount.

Advancing subsurface analysis: Integrating computer vision and deep learning for the near real-time interpretation of borehole image logs in the Illinois Basin-Decatur Project

The accurate quantification and mapping of subsurface natural fracture systems using borehole imaging logs are critical for the success of CO2 sequestration in geologic formations, optimization of engineered geothermal systems, and hydrocarbon production enhancement. However, traditional interpretation processes suffer from time-consuming procedures and human bias. To address these challenges and expedite fracture analysis, we investigated the application of integrated computer vision and DL workflows to automate image log analysis. Specifically, the design of our workflow was crafted to swiftly detect fractures and baffles by using actual electrical resistivity of borehole wall from microresistivity imaging device alongside their binary representation. This novel approach significantly reduces computational time while providing invaluable insights. By incorporating conventional logging and microseismic data, we present a regional subsurface natural fracture mapping technique. Through the minimization of human bias in image log analysis, our automated workflow achieves reduced fracture interpretation time and costs while ensuring robust and reproducible results. We demonstrated the efficacy of our approach by applying the workflow to the Illinois Basin-Decatur Project site. The automated workflow successfully identified major fractured zones, multiple baffles, and an interbedded layer with a high resolution of 0.01 ft or 0.12 in. (0.3 cm) and can be upscaled to any desired resolution. Validation through microseismic and image log interpretations allows for accurate and near-real-time mapping of fractures and baffles, significantly enhancing CO2 pressure forecasting and postinjection site care. Our approach stands out due to its robustness, consistency, and reduced computational cost compared with alternative feature extraction technologies. It presents exciting possibilities for advancing CO2 sequestration and engineered geothermal efforts by offering comprehensive and efficient fracture mapping solutions. This technology can contribute significantly to the optimization of CO2 sequestration projects, facilitating sustainable environmental practices, and combating climate change.

Digitalization of toxicology: improving preclinical to clinical translation.

Though the portfolio of medicines that are extending and improving the lives of patients continues to grow, drug discovery and development remains a challenging business on its best day. Safety liabilities are a significant contributor to development attrition where the costliest liabilities to both drug developers and patients emerge in late development or post-marketing. Animal studies are an important and influential contributor to the current drug discovery and development paradigm intending to provide evidence that a novel drug candidate can be used safely and effectively in human volunteers and patients. However, translational gaps-such as toxicity in patients not predicted by animal studies-have prompted efforts to improve their effectiveness, especially in safety assessment. More holistic monitoring and "digitalization" of animal studies has the potential to enrich study outcomes leading to datasets that are more computationally accessible, translationally relevant, replicable, and technically efficient. Continuous monitoring of animal behavior and physiology enables longitudinal assessment of drug effects, detection of effects during the animal's sleep and wake cycles and the opportunity to detect health or welfare events earlier. Automated measures can also mitigate human biases and reduce subjectivity. Reinventing a conservative, standardized, and traditional paradigm like drug safety assessment requires the collaboration and contributions of a broad and multi-disciplinary stakeholder group. In this perspective, we review the current state of the field and discuss opportunities to improve current approaches by more fully leveraging the power of sensor technologies, artificial intelligence (AI), and animal behavior in a home cage environment.

Biomechanical and histomorphometric analysis of osseodensification drilling versus conventional technique: a systematic review and meta-analysis.

This systematic review aimed to search the literature for the answer to the following questions. In human studies: Does the osseodensification technique increase the resonance frequency analysis given in implant stability quotient value and the insertion torque value compared to the conventional technique? In animal studies: Does the osseodensification technique increase implant stability quotient, bone-to-implant contact, and bone area fraction occupancy values over the conventional technique? A search for studies was carried out in eight databases until August 2021. Out of the 447 publications found, 11 were included. In human studies, osseodensification technique showed better results for implant stability quotient values with a summarized median difference of 8.57. As for secondary stability, there was no significant difference, with summarized median difference of 4.49 in favor of the osseodensification technique. In animal studies, all results were favorable to the osseodensification technique. Regarding insertion torque, bone-to-implant contact, and bone area fraction occupancy between counterclockwise osseodensification technique vs conventional, the mean difference was 46.79 for insertion torque, 2.17 for bone-to-implant contact, and 2.11 for bone area fraction occupancy. High heterogeneity was observed between the studies. The risk of bias in humans was moderate in three studies and low in one; and in animal studies, four presented moderate risk, two low risk, and one high risk. The certainty of evidence ranged from low to moderate. The osseodensification technique showed improvement concerning the resonance frequency and the insertion torque value of implants in human studies. In addition, it increased the values of bone-to-implant contact, bone area fraction occupancy, and implant stability quotient in animal studies, when compared to the conventional technique.

An inferential spatiotemporal approach for knowledge synthesis to identify trends in public health research

BackgroundDecisions follow patterns that are introduced by human perception. Research and development (R&D) are influenced by these patterns. Furthermore, R&D publications can represent repetitive attempts to solve similar, or the same problems. Literature reviews serve as an important tool for identifying these trends, but they are time consuming. The time commitment of a literature review can be reduced by using a sample of research. This will allow an infinite population of research to be generalized. Additionally, spatiotemporal analysis is most appropriate for fields that follow time and geographic trends, such as public health. Also, using research locations to perform this analysis potentially captures the social return of R&D, as knowledge gained. As a result, an inferential spatiotemporal methodological framework is introduced to quickly identify research trends using public health research. This was applied to a childhood Pb exposure case study. MethodsA body of more than 1000 childhood elevated blood lead (Pb) level (EBLL) research articles were used to extract publication years, research locations, and subtopics. These publications were grouped into research locations (i.e., U.S. states where research was conducted; not publication location) and averaged over years published (i.e., 29 years). Binary indicator variables were derived using the subtopics extracted and the periods identified in time trend analyses. Explanatory variables were used to conduct hypothesis testing. Significant variables were used to generalize the population of the annual average EBLL articles written per state. ResultsThe range of the annual average of EBLL research articles by state was 0–1.7 articles, with a mean of 0.3 articles. Thirty-eight explanatory variables suggested a significant effect on research article production. These included temporal, sociodemographic, education, structure age, environmental, and economic variables. The strongest effect on research production for U.S. states came from the number of structures built before 1950. A predictive model was selected to generalize the population of articles using time-periods 1990-95, environmental subtopic, and structures built before 1950. The locations with the most research production for this topic were California and New York. The locations with the least research production for this topic were Alaska, Hawaii, Nevada, Wyoming, North Dakota, South Dakota, Mississippi, Delaware, and New Hampshire. ConclusionIf the trend for R&D is to make fast decisions, more human bias will be introduced into the decision-making process. Analytical tools that enable researchers to identify trends and ask more questions about their field will mitigate these biases. This hypothesis testing and predictive modeling methodology provide researchers and other decision makers with analytical tools they can use to quickly identify research trends and narrow their field of research. Additionally, this analysis potentially captures the impact of discovered ideas, as a social return spillover, for this topic.

MagiCapture: High-Resolution Multi-Concept Portrait Customization

Large-scale text-to-image models including Stable Diffusion are capable of generating high-fidelity photorealistic portrait images. There is an active research area dedicated to personalizing these models, aiming to synthesize specific subjects or styles using provided sets of reference images. However, despite the plausible results from these personalization methods, they tend to produce images that often fall short of realism and are not yet on a commercially viable level. This is particularly noticeable in portrait image generation, where any unnatural artifact in human faces is easily discernible due to our inherent human bias. To address this, we introduce MagiCapture, a personalization method for integrating subject and style concepts to generate high-resolution portrait images using just a few subject and style references. For instance, given a handful of random selfies, our fine-tuned model can generate high-quality portrait images in specific styles, such as passport or profile photos. The main challenge with this task is the absence of ground truth for the composed concepts, leading to a reduction in the quality of the final output and an identity shift of the source subject. To address these issues, we present a novel Attention Refocusing loss coupled with auxiliary priors, both of which facilitate robust learning within this weakly supervised learning setting. Our pipeline also includes additional post-processing steps to ensure the creation of highly realistic outputs. MagiCapture outperforms other baselines in both quantitative and qualitative evaluations and can also be generalized to other non-human objects.

Contrastive regularization guided label refurbishment for fault diagnosis under label noise

Conventional intelligent fault diagnosis studies require precise labels by default. However, label noise is inevitable in real industrial environments due to human errors, measurement biases, data transmission and storage errors. The performance of existing data-driven methods may be severely affected by mislabels. In this paper, we investigate the fault diagnosis issue of rotating machinery under noisy labels. A generalized framework contrastive regularization guided label refurbishment (CRLR) is proposed, which utilizes contrastive regularization function extracting class prototype to guide label refurbishment. The learned contrastive representations can suppress the misleading effect of noisy labels. The theoretical analysis from an information theoretic perspective also ensures the effectiveness of CRLR. Validation experiments have been conducted on public and private collected datasets using the ResNet-50 as the backbone network. The experiments demonstrate that our proposed framework outperforms the current state-of-the-art methods at different noise rates.

Real-Time Automated Geosteering Interpretation Combining Log Interpretation and 3D Horizon Tracking

Existing methods for estimating formation boundaries from well-log data only analyze the formation along the wellbore, failing to capture changes in the 3D formation structure around it. This paper presents a method for real-time 3D formation boundary interpretation using readily available well logs and seismic image data. In the proposed workflow, the mean formation boundary is estimated as a curve following the well path. 3D surfaces are then fitted through this boundary curve, aligning with the slopes and features in the seismic image data. The proposed method is tested on both synthetic and field datasets and illustrates the capabilities of accurate boundary estimation near the well path and precise representation of boundary shape changes further away from the well trajectory. With this fully automated geological interpretation workflow, human bias and interpretation uncertainty can be minimized. Subsurface conditions can be continually updated while drilling to optimize drilling decisions and further automate the geosteering process.

Estimation of Temperature and Salinity from Marine Seismic Data—A Two-Step Approach

Ocean-water temperature and salinity are two vital properties that are required for weather-, climate-, and marine biology-related research. These properties are usually measured using disposable instruments at sparse locations, typically from tens to hundreds of kilometers apart. Laterally interpolating these sparse measurements provides smooth temperature and salinity distributions within the oceans, although they may not be very accurate. Marine seismic data, on the other hand, show visible reflections within the water-column which are primarily controlled by subtle sound-speed variations. Because these variations are functions of the temperature, salinity, and pressure, estimating sound-speed from marine seismic data and relating them to temperature and salinity have been attempted in the past. These seismically derived properties are of much higher lateral resolution (less than 25 m) than the sparse measurements and can be potentially used for climate and marine biology research. Estimating sound-speeds from seismic data, however, requires running iterative seismic inversions, which need a good initial model. Currently practiced ways to generate this initial model are computationally challenging, labor-intensive, and subject to human error and bias. In this research, we outline an automated method to generate the initial model which is neither computational and labor-intensive nor prone to human errors and biases. We also use a two-step process of, first, estimating the sound-speed from seismic inversion data and then estimating the salinity and temperature. Furthermore, by applying this method to real seismic data, we demonstrate the feasibility of our approach and discuss how the use of machine learning can further improve the computational efficiency of the method and make an impact on the future of climate modeling, weather prediction, and marine biology research.