Mixed Data Research Articles

Gentrification and touristification in Lavapiés: Causes and consequences of a unique interplay in today’s Southern Europe

ABSTRACT This article aims to examine the current interplay between gentrification and extractivist touristification in the historical neighborhood of Lavapiés in Madrid (Spain), which is quite singular in the southern European Spanish urban context. After discussing the often-complex coexistence between gentrification and touristification, this article examines the current state of such interplay in Lavapiés through mixed methods research, combining ethnography and data collection from official statistical sources. The article concludes by suggesting the importance of considering local political contexts in the study of gentrification and touristification in order to avoid overly general comparisons between cities that could lead to a misinterpretation of the transnational scale of the extractivist touristification in central urban areas at a European scale (and beyond).

Agro‐Morphological Variability Among Sweetpotato [Ipomoea batatas (L.) Lam] Clones Collected From Niger and Nigeria

ABSTRACTSweetpotato is gaining momentum in Niger and Nigeria due to its contribution to food and nutritional security. The aim of this study was to assess the variability among the sweetpotato clones collected from Niger and Nigeria. Hundred sweetpotato clones were evaluated during the rainy season 2020–2021 using alpha lattice design. Data were collected on quantitative and qualitative traits and analysed using R software. The Analysis of variance exhibited a highly significant difference (p < 0.001) among clones for all traits, indicating high variability among clones. The study revealed high values of genotypic and phenotypic coefficients of variation in all the traits except for mature leaf size and storage root length, which indicated the presence of wide gene variability. High heritability along with a high percentage of genetic advances suggests possible and rapid improvement of these traits. The results of factorial analysis of mixed data grouped the hundred clones into three main clusters, C1, C2 and C3 with 49, 14 and 37 clones, respectively. Clones EBO/SP8, JOHN and TIS‐81‐64 × 1 had the highest storage root yield, whereas clones Nwa Oyorima, PYT/25/19 and PYT/4/19 × 1 had the best in upper biomass yield; therefore, these clones could be further used as potential sources to improve these traits.

Evaluating the Dynamic Interrelations Among Coping Self-Efficacy, Coping Behaviors, and PTSD Symptoms: Analysis Using Mixed Methods EMA Data

ABSTRACT Stress coping models posit that traumatic recovery is a dynamic process where ongoing coping self-efficacy (CSE) appraisals guide selection and utilization of coping strategies. The success of these coping strategies in turn influence future CSE appraisals and coping efforts, ultimately leading to trauma (non) recovery. However, there is very limited existing research modeling these dynamic interrelations, particularly data utilizing assessments that can capture these dynamics as they unfold. The current study utilized ecological momentary assessment (EMA) data from 85 highly exposed rural hurricane survivors to evaluate these recovery dynamics. Specifically, previous day CSE was evaluated as a predictor of the type of coping (approach, avoidance, religious) utilized in response to a stressor. Further, the type of coping utilized in response to a stressor was evaluated as a predictor of next day PTSD symptoms and CSE. Pre-stressor PTSD symptoms were included in the models as a control variable. Results supported that higher pre-stressor CSE was associated with a higher likelihood of using approach coping the following day (OR = 1.07), whereas lower pre-stressor CSE was associated with a lower likelihood of using avoidance (Inverse OR = 1.09) and religious (Inverse OR = 1.06) coping the following day. Use of avoidance coping predicted lower next day PTSD (β = -.23) and CSE (β = -.14). Results overall supported the dynamic interrelations among trauma symptomology, CSE, and coping behaviors. Future work should utilize intensive time series data to model these interrelations over time in the trauma recovery process.

CA2CL: Cluster-Aware Adversarial Contrastive Learning for Pathological Image Analysis.

Pathological diagnosis assists in saving human lives, but such models are annotation hungry and pathological images are notably expensive to annotate. Contrastive learning could be a promising solution that relies only on the unlabeled training data to generate informative representations. However, the majority of current methods in contrastive learning have the following two issues: (1) positive samples produced through random augmentation are less challenging, and (2) false negative pairs problem caused by negative sampling bias. To alleviate the above issues, we propose a novel contrastive learning method called Cluster-Aware Adversarial Contrastive Learning (CA2CL). Specifically, a mixed data augmentation technique is provided to learn more transferable representations by generating more discriminative sample pairs. Furthermore, to mitigate the effects of inherent false negative pairs, we adopt a cluster-aware loss to identify similarities between instances and incorporate them into the process of contrastive learning. Finally, we generate challenging contrastive data pairs by adversarial learning, and adversarially learn robust representations in the representation space without the labeled training data, which aims to maximize the similarity between the augmented sample and the related adversarial sample. Our proposed CA2CL is evaluated on two public datasets: NCT-CRC-HE and PCam for the fine-tuning and linear evaluation tasks and on two other public datasets: GlaS and CARG for the detection and segmentation tasks, respectively. Extensive experimental results demonstrate the superior performance improvement of our method over several Self-supervised learning (SSL) methods and ImageNet pretraining particularly in scenarios with limited data availability for all four tasks. The code and the pre-trained weights are available at https://github.com/junjianli106/CA2CL.

Using Mixed Methods to Evaluate Risk Minimisation Programs in Europe and the USA: An Innovative Blueprint.

Significant methodological shortcomings have been documented to date in risk minimisation program evaluations for medicinal products, including overreliance on survey methods alone. Recently updated guidances from the Food and Drug Administration (FDA) and the European Medicines Agency (EMA) recommend the use of frameworks and mixed methods designs to improve the rigor of these assessments. The purpose of this paper was to exemplify how a mixed methods approach, guided by an implementation science framework, can be used to design the evaluation of a risk minimisation program. We selected the Practical, Robust, Implementation and Sustainability Model (PRISM) as the implementation science framework to guide our mixed methods approach. PRISM provides a comprehensive and systematic approach to measuring the key domains relevant to the implementation and outcomes of a risk minimisation program. We mapped the PRISM domains to the evaluation dimensions described in the EMA and FDA guidances. We then specified a mixed methods evaluation design and data collection methods using a fictitious risk minimisation program as a case study for illustrative purposes. On the basis of our case study, we developed quantitative and qualitative measures, including specific items for surveys and interviews, for both formative and summative evaluations. For both the formative and summative evaluations, measures focussed on assessing (1) contextual factors that could affect program implementation and impact and (2) outcomes including implementability and acceptability as well as degree of program reach, adoption, implementation, effectiveness and maintenance. Mixed methods, guided by a well-established implementation science framework, can be applied to ensure comprehensive formative and summative evaluations that provide fit-for-purpose information that may inform regulatory decision-making.

Mixed frequency data and portfolio selection: A novel approach integrating DEA with mixed frequency data sources

Mixed frequency data and portfolio selection: A novel approach integrating DEA with mixed frequency data sources

Joint Mixing Data Augmentation for Skeleton-Based Action Recognition

Skeleton-based action recognition is beneficial for understanding human behavior in videos, and thus has received much attention in recent years as an important research area in action recognition. Current research focuses on designing more advanced algorithms to better extract spatio-temporal information from skeleton data. However, due to the small amount of data in the existing skeleton dataset and the lack of effective data augmentation methods, it is easy to lead to overfitting in model training. To address this challenge, we propose a mix-based data augmentation method, Joint Mixing Data Augmentation (JMDA), which can generally improve the effectiveness and robustness of various skeleton-based action recognition algorithms. In terms of spatial information, we introduce SpatialMix (SM), a method that projects the original 3D skeleton discrete information into a 2D space. Then, SM mixes the projected spatial information between two random samples during the training process to achieve the spatial-based mixing data augmentation. Concerning temporal information, we propose TemporalMix (TM). Leveraging the temporal continuity in skeleton data, we perform a temporal resize operation on the original skeleton data, and then merge two random samples during training to achieve the temporal-based mixed data augmentation. Additionally, we analyze the Feature Mismatch (FM) problem caused by introducing mix-based data augmentation into skeleton data. Then we propose a new data preprocessing method called Feature Alignment (FA) to effectively address this problem and improve model performance. Moreover, we propose a novel training pipeline, Joint Training Strategy (JTS), which combines multiple mix-based data augmentation methods for further improvement of model performance. Specifically, our proposed JMDA is plug-and-play and widely applicable to skeleton-based action recognition models. At the same time, the application of JMDA does not increase the model parameters and there is almost no additional training cost. We conduct extensive experiments on NTU RGB+D 60 and NTU RGB+D 120 datasets to demonstrate the effectiveness and robustness of the proposed JMDA on several mainstream skeleton-based action recognition algorithms.

"I still can't forget those words": mixed methods study of the persisting impact on patients reporting psychosomatic and psychiatric misdiagnoses.

This research aimed to improve understanding of persisting impacts of patient-reported psychosomatic and psychiatric misdiagnoses on patients with systemic autoimmune rheumatic diseases (SARDs). Mixed methods data from two SARDs cohorts were analysed (N = 1,543 and N = 1,853). Validated instruments and patient-designed questions were used to measure self-reported depression, anxiety and mental wellbeing, in addition to medical relationships and healthcare behaviours. Comparative tests were used to evaluate differences between patients reporting a psychosomatic and/or psychiatric misdiagnoses and other patients. Persisting adverse outcomes of perceived psychosomatic and psychiatric misdiagnoses were identified in multiple domains. This included >80% of patients reporting that it had damaged their self-worth, and 72% reporting that it still upset them. Patients reporting psychosomatic and/or psychiatric misdiagnoses had significantly lower mental wellbeing, and higher depression and anxiety levels (all p< 0.001), and lower levels of satisfaction with every aspect of medical care, compared with patients reporting no psychosomatic or psychiatric misdiagnoses. Psychosomatic and psychiatric misdiagnoses had varying associations with healthcare behaviours, including a significantly higher likelihood of under-reporting symptoms (p< 0.001) and healthcare avoidance (p= 0.012), but not with medication adherence (p= 0.2). Thematic analysis of qualitative data revealed that symptom under-reporting and healthcare avoidance often resulted from distrust and fear that symptoms would be disbelieved and misattributed again. Patient-reported psychosomatic and psychiatric (mis)diagnoses are associated with persisting adverse impacts in multiple domains including mental health, medical relationships, self-worth, and some healthcare behaviours. Health services and clinicians should consider these potential adverse impacts on patients and offer support to reduce any persisting negative impacts.

Advancing a relational approach through online education: A mixed methods assessment in older adults' care.

To evaluate the experience of health-care workers who completed an online professional development program designed to support family carergivers in care teams in personal care homes. The evaluation focussed on issues of access and uptake of education modules. An explanatory sequential mixed methods design was used to include data from posteducation surveys, and online module analytics of use, to inform interview questions. Demographic data, survey data and video analytics were described. Focus group data were analysed using reflexive thematic analysis. Mixed method data integration enabled inference development. Across eight personal care homes located in Canada and Australia, of 114 health-care workers (mainly care aides) who commenced the online education modules, 89 watched and completed the postassessment. Most agreed that the educational content increased their understanding, and the videos increased their knowledge. Ten participants took part in focus groups interviews. The education was perceived as another way to learn and was considered beneficial for individual and team practice. Health-care workers found it difficult to find time to undertake learning opportunities outside work when their lives were busy and involved other responsibilities or multiple jobs. The need for organisational support to coordinate time during work to undertake work-related learning was highlighted. Caregiver-centred care education exemplifies enhancing relational care among health-care staff and family caregivers, beneficial for individual and team practice. By valuing education for health-care workers through creating opportunities for staff to engage during work time, a cultural environment that embodies continuous learning can be created in sectors caring with older adults.

Predictive Model to Analyse Real and Synthetic Data for Learners' Performance Prediction Using Regression Techniques

Predicting learner performance with precision is critical within educational systems, offering a basis for tailored interventions and instruction. The advent of big data analytics presents an opportunity to employ Machine Learning (ML) techniques to this end. Real-world data availability is often hampered by privacy concerns, prompting a shift towards synthetic data generation. This study presents an empirical comparison of real, synthetic, and mixed (real + synthetic) data sets in forecasting learner performance, deploying an array of regression-based ML algorithms, including Random Forest, Gradient Boosting, XG Boost, K-nearest Neighbor, and Support Vector Regression. Our methodology encompasses the generation of synthetic data via generative model, followed by the application of these algorithms to each data set. The models are evaluated using precision metrics to assess their predictive accuracy. The study unveils that synthetic data can rival real data in predictive capabilities, with combined data sets achieving up to 87.76% accuracy, underscoring the efficacy of hybrid data approaches. These insights advocate for the integration of synthetic data as a practical substitute in scenarios with limited access to real data, fostering advancements in educational technology and ML.

Multivariate Approach to Antimicrobial Residue Concentrations in Animal-Derived Products: An Analytical Review.

Antimicrobial resistance (AMR) represents an alarming global public health concern exacerbated by livestock antibiotic misuse, affecting humans and the environment. However, the precise magnitude of antimicrobial residue concentrations in animal-derived products remains not well understood. This study aimed to quantify antimicrobial residues in animal products through an analytical literature review. This review covered the scientific articles from 1977 to 2020. The antimicrobials were classified according to the European Medicines Agency (EMA) guidelines into four categories. The final database comprised seven qualitative variables (antibiotic, antibiotic class, region, country, decade, EMA category, animal product and animal species) and one quantitative variable (residue concentration recorded as µg/kg). Due to the number of variables involved in the study, a multivariate analysis approach was used using a Factor Analysis of Mixed Data (FAMD) carried out in R. The highest concentrations of antimicrobial residues were detected in fish samples, followed by egg. Notably, concentrations of ruminant-derived products were lower than to monogastric. β-Lactam was the most prevalent residue followed by aminoglycosides, sulphonamides and quinolones, respectively. Moreover, South America had the highest residues levels, followed by Asia and Europe. The multivariate analysis reveals a possible association between the EMA category, animal species, antimicrobial class and animal product. In conclusion, the concentration of antimicrobial residues in products of animal origin depends mainly on their origin (product, species and geographic region), showing the highest concentrations in products derived from fish and poultry.

A comprehensive survey on recent feature selection methods for mixed data: Challenges, solutions and future directions

A comprehensive survey on recent feature selection methods for mixed data: Challenges, solutions and future directions

Community as Medicine: A Novel Approach to Improve Health Behaviors and Mental Well-Being for Vulnerable Populations.

Evaluate a group-based, trauma-informed health coaching model based on Lifestyle Medicine pillars, delivered by community health coaches, to improve health behaviors and mental well-being for vulnerable populations. This program evaluation analyzed quantitative longitudinal data with linear mixed models and qualitative data with reflexive thematic analysis. Participants (n = 720) were low-income adults referred through Federally Qualified Health Centers (FQHCs), who participated in weekly 90-120 minute groups for 3 months. Data were collected via monthly surveys, including the PHQ-9, GAD-7, UCLA 3-item loneliness, exercise as a vital sign, a 2-item dietary screener, and 3 qualitative questions. Participants saw significant reductions in depression, anxiety, and isolation, and significant increases in daily servings of fruits and vegetables, and weekly minutes of exercise. The qualitative analysis identified 4 themes related to the drivers and reinforcers of positive behavior change and improved mental well-being. Initial data suggest this model helps reduce depression, anxiety, and isolation, and promotes positive behavior change within populations most impacted by health inequity. Qualitative results identified drivers of positive change, such as creating a sense of belonging and mutual support. Future research should consider how to continue to scale this program to a variety of populations and across different settings.

Species delimitation in the Polygonatum odoratum complex (Asparagaceae) based on phylogenomic and morphometric data

AbstractDefining species and delineating boundaries among closely related species is challenging particularly for species complexes with a recent history of diversification. The Polygonatum odoratum complex, consisting of seven taxa, presents a compelling example showing unresolved taxonomic boundaries. Here, we aim to delimit taxonomic boundaries in the complex using genomic and morphometric data. Target enrichment was employed to analyze 40 accessions representing the seven taxa in the complex. We inferred the species tree using coalescent and maximum likelihood approaches with nuclear and plastid data. Additionally, we examined 14 morphological characters from 851 herbarium specimens and conducted morphometric analysis using factor analysis of mixed data. The phylogenomic analysis, based on 297 genes, strongly supports the presence of three monophyletic clades with a recent diversification history (&lt;4 million years). Coupled with our morphological examinations, our findings demonstrate the existence of three distinct species, revising the previous classification of seven taxa. By applying an integrated species concept, our study resolves the species boundaries and updates the taxonomy of the P. odoratum complex.

Code Mixing Found in the Leonardo’s Podcast with Satria Mahatir

Nowadays, the use of two or more languages social media is a common phenomenon. This study aims to find out the types of Code Mixing in The Leonardo’s Podcast with Satria Mahatir and to analyze the reason of the usage of code mixing on The Leonardo's Podcast with Satria Mahatir. The data source of this study is The Leonardo's podcast with Satria Mahatir. This research method includes descriptive qualitative by means of observation, watching, and writing the utterances with code mixing data. The theory used in analyzing the data is according to Muysken (2000). The results of the findings, found that the type of code mixing are insertion, alternation, and congruent lexicalization. In the research carried out by the researcher, the number of code mixing found were 44 data which include insertion code mixing 40 data, alternation code mixing 3 data, and congruent lexicalization only 1 data. Insertion is used dominantly in the data source due to the lack of word equivalents and the influence of local prestige culture that can cause this to happen.

Mobilizing Health Promotion Through Canada's Student Mental Health Network: Concurrent, Mixed Methods Process Evaluation.

Mental health issues among Canadian postsecondary students are prevalent. In tandem, an increased acknowledgment of the need for upstream mental health support has been highlighted. While the majority of institutions offer some form of mental health promotion, research suggests students are failing to access support due to barriers including lack of awareness, geographical and financial barriers, and lack of relevance in offerings. Canada's Student Mental Health Network is a web-based knowledge mobilization initiative designed to fill these gaps. With content created and curated "for-students, by-students" and reviewed by subject matter experts, the Network serves as a one-stop shop for evidence-based, mental health support for postsecondary students. The goal of this research was to conduct the first component of a comprehensive program evaluation of the Network. This paper details a formative, process evaluation after approximately 1 year of operations, with the goal of assessing acceptability and feasibility. Using a concurrent mixed methods study design, quantitative and qualitative data were simultaneously collected from students in order to evaluate the acceptability and feasibility of the Network as a mental health promotion resource. Quantitative data were automatically collected through Google Analytics via the website over the course of the first year of operations. Data collected included the number of users accessing the website, user engagement, and user "stickiness." Quantitative data were used to evaluate both accessibility and feasibility. Qualitative data were collected via individual, digital interviews conducted with a modest sample of students (n=8) across areas and levels of study. Qualitative data derived more detailed insights into user experience and website attributes, as well as feedback on content delivery, providing evidence used to evaluate feasibility. A total of 1200 users globally accessed the Network within the first year of operations, with Canadian users accounting for nearly 90% of total website traffic. An overall 66% engagement rate was observed, with the average user visiting 7 pages per session. Further support for the acceptability of the Network is demonstrated in the Canada-wide reach of the content development and review team. Evidence for the feasibility of the Network was observed through website use statistics indicating the most frequently viewed pages aligned with our goals: providing mental health education and increasing awareness of available resources. Qualitative feedback provided additional context surrounding the feasibility of the space, including positive feedback on the esthetics, relevance, usability, inclusion, and accessibility. Areas for content expansion and improvements to accessibility were also identified. The results of this study provide evidence in support of the feasibility and acceptability of the Network as a web-based knowledge mobilization initiative in support of postsecondary students' mental health and well-being. Future research will pursue a summative, impact assessment to evaluate utility.

Population-level transitions in observed difficulties through childhood and adolescence.

In an attempt to better characterize the complexity of difficulties observed within developing populations, numerous data-driven techniques have been applied to large mixed data sets. However, many have failed to incorporate the core role of developmental time in these approaches, that is, the typical course of change in behavioral features that occurs over childhood to adolescence. In this study, we utilized manifold projections alongside a gradient-boosting model on data collected from the Millennium Cohort Study to unpack the central role of developmental time in how behavioral difficulties transition between the ages of 5, 11, and 17. Our analysis highlights numerous observations: (a) Girls develop relatively greater internalized behavioral problems during adolescence; (b) in the case of a chaotic home environment, co-occurring internalizing and externalizing difficulties tend to persist during childhood; (c) peer problems were the most likely to persist over the whole 12-year period (especially in the presence of early maternal depression and poor family relationships); and (d) there were two pathways with distinct risk factors leading to antisocial behaviors in adolescence-an early-childhood onset pathway and later adolescent onset pathway. Our findings provide evidence that investigations of child and adolescent difficulties must be open to the possibility of multiple subgroups and variability in trajectory over time. We further highlight the crucial role of family and social support and school experience-related factors in predicting children's outcomes. (PsycInfo Database Record (c) 2025 APA, all rights reserved).

Distinct Aerosol Impacts on Local Scale Convective Rainfall Between Sichuan Basin and North China Plain Regions in China

AbstractAerosols modulate cloud and precipitation processes through complex aerosol‐radiation interactions (ARI) and aerosol‐cloud interactions (ACI). The influence of aerosols on precipitation varies regionally due to many factors, including aerosol characteristics, precipitation types, and meteorological conditions. Using high‐resolution precipitation data from the hourly China merged precipitation analysis (CMPA) Version 1.0 and aerosol mixing ratio data from the Modern‐Era Retrospective Analysis for Research and Application Version 2 (MERRA‐2) during the warm seasons of 2015–2020, this study employs sulfate mixing ratio () as a CCN proxy to investigate the distinct impacts of aerosols on warm‐topped and cold‐topped rainfall across the Sichuan Basin (SCB) and the North China Plain (NCP) regions in China. Our findings reveal that in SCB, aerosols suppress warm‐topped rain by stabilizing the atmosphere, reducing the cloud effective radius (re) and potentially weakening collision‐coalescence processes. Conversely, in cold‐topped rain events, aerosols initially enhance rainfall, but subsequently suppress it at higher concentrations. In NCP, suspended dust increases with rising acting as both giant cloud condensation (GCCNs) and ice nuclei (INs). This enhances cloud re and promotes rainfall across both cloud types. Furthermore, the aerosol radiative effect destabilizes the atmosphere, amplifying these processes. Despite variations in meteorological conditions, aerosols consistently exert a significant influence on rainfall. This research highlights the distinct aerosol impacts on rainfall in the two regions and underscores the need for further study to improve climate predictions.

PharaCon: a new framework for identifying bacteriophages via conditional representation learning.

Identifying bacteriophages (phages) within metagenomic sequences is essential for understanding microbial community dynamics. Transformer-based foundation models have been successfully employed to address various biological challenges. However, these models are typically pre-trained with self-supervised tasks that do not consider label variance in the pre-training data. This presents a challenge for phage identification as pre-training on mixed bacterial and phage data may lead to information bias due to the imbalance between bacterial and phage samples. To overcome this limitation, we proposed a novel conditional BERT framework that incorporates label classes as special tokens during pre-training. Specifically, our conditional BERT model attaches labels directly during tokenization, introducing label constraints into the model's input. Additionally, we introduced a new fine-tuning scheme that enables the conditional BERT to be effectively utilized for classification tasks. This framework allows the BERT model to acquire label-specific contextual representations from mixed sequence data during pre-training and applies the conditional BERT as a classifier during fine-tuning, and we named the fine-tuned model as PharaCon. We evaluated PharaCon against several existing methods on both simulated sequence datasets and real metagenomic contig datasets. The results demonstrate PharaCon's effectiveness and efficiency in phage identification, highlighting the advantages of incorporating label information during both pre-training and fine-tuning. The source code and associated data can be accessed at https://github.com/Celestial-Bai/PharaCon.

Geopolymer Concrete Optimization: A Critical Review of Supplementary Materials and Performance Enhancement

In the realm of environmentally friendly building materials, geopolymer concrete stands out as a viable substitute for traditional concrete made of cement. This literature review explores recent developments and innovations in geopolymer concrete, with an emphasis on optimizing its properties through the use of various supplementary materials. The aim is to enhance both the mechanical strength and geopolymer concrete's microstructural properties, while also investigating the cost efficiency of these enhancements. The review systematically examines different combinations of supplementary cementitious materials, analyzing their impact on mechanical stability and long-term performance. Additionally, the study delves into the creation of precise Artificial Neural Network models using mixed synthetic data to accurately predict geopolymer concrete properties. Microstructural properties of geopolymer concrete are evaluated through XRD and SEM analyses, providing valuable insights into its structural integrity. Furthermore, a comparative analysis of the cost parameters between traditional concrete and geopolymer concrete is conducted, shedding light on the economic viability of GPC in construction projects. This thorough evaluation aims to support ongoing research and development in the field of geopolymer concrete, offering valuable insights for engineers, researchers, and industry professionals seeking sustainable and efficient construction materials. Incorporating supplementary cementitious materials, such as fly ash, slag, and metakaolin, significantly enhances the mechanical strength and durability of geopolymer concrete (GPC), providing superior resistance to environmental degradation, such as sulfate attack, compared to traditional Portland cement concrete. The application of artificial intelligence, specifically ANN modeling with mixed synthetic datasets, demonstrates a high degree of accuracy in predicting the compressive strength and long-term performance of GPC. This facilitates the development of optimized and reliable mix designs tailored to diverse construction needs. A detailed cost analysis highlights the economic and environmental viability of GPC. While its initial production costs may be higher due to specialized materials, its long-term cost-effectiveness is evident through enhanced durability and reduced maintenance. Its lower carbon footprint makes GPC a sustainable alternative for future projects while using industrial by-products promotes a circular economy. Major Findings: Geopolymer Concrete (GPC) can significantly enhance mechanical strength and microstructural properties through various supplementary cementitious materials. Advanced techniques like Artificial Neural Networks (ANN) and microstructural analyses (XRD, SEM) provide deeper insights into GPC's stability and performance. Additionally, GPC demonstrates economic viability compared to traditional concrete, offering sustainable benefits for construction projects.