Termez State University

Hyperspectral imaging (HSI) enables real-time, non-invasive tissue characterization by capturing biochemical signatures across hundreds of spectral bands simultaneously. However, the computational demands of processing vast hyperspectral datasets create significant bottlenecks for clinical deployment. This chapter examines how edge computing and hardware acceleration technologies—including Field-Programmable Gate Arrays (FPGAs), Graphics Processing Units (GPUs), and specialized neural accelerators—enable real-time HSI processing at the point of care. We analyze architectural trade-offs, implementation strategies, and clinical workflow integration approaches that reduce processing times from minutes to milliseconds. Recent advances in embedded platforms achieve throughput rates exceeding 250 MSamples/s while maintaining power consumption below 15W, making intraoperative HSI guidance practically achievable.

This study aimed to design and evaluate a proactive, school-based educational framework functioning as an early diversion mechanism for vulnerable youth at risk of future justice system involvement. The study addressed the question of whether the presence of structured early identification, individual learning support, and combined behavioral and social-emotional interventions might have led to a decrease in the occurrence of academic disengagement, disciplinary escalation, and criminogenic risk indicators. It adopted a quasi-experimental longitudinal research design where 180 at-risk students (n = 90 intervention; n = 90 control) were used. These groups were tested to be equivalent at the baseline (p > 0.05). The intervention was provided in a 12-month multi-tiered support model that incorporated academic support, behaviour regulation, restorative practices, and involvement of family. Repeated-measures ANOVA, multilevel modeling, multilevel mediation analysis, and thematic qualitative analysis were used to assess the outcomes. There was significant time × group interaction effects on GPA (F (1, 178) = 18.42, p < 0.001) and attendance (F (1, 178) = 21.75, p < 0.001) having large effect sizes (d = 0.82; d = 0.91). There were significant decreases in disciplinary referrals and behavioral risk scores in behavioral outcomes, with intervention attendance being a significant predictor of behavioral risk (β = −14.72, p < 0.001). The risk indicators on justice decreased greatly (t (178) = 5.63, p < 0.001), and the probability of high-risk classification reduced from 62% to 29% of high-risk in the intervention group. The mediation analysis showed a significant indirect effect (−6.45, 95% CI [−9.12, -3.78]) when it comes to academic engagement and self-regulation. For vulnerable adolescents, proactive, composite educational innovations can generate positive academic results and significantly diminish danger in terms of conduct and equity, and place schools as efficient early diversionary frameworks.

The aim of the study was to estimate the efficiency of a peer-based teaching and cooperative learning intervention in forensic users with a complex neurodivergent profile (Autism Spectrum Disorder, Attention-Deficit/Hyperactivity Disorder, and intellectual disabilities). The research question was that peer-facilitated learning would increase social communication, emotional regulation, engagement, and self-efficacy and decrease behavioral incidents as compared to usual clinician-facilitated interventions. The quasi-experimental mixed-methods study design was used, including 84 adult forensic service users who were divided into a collaborative learning group (led by peers, n=42) and a control group (led by a clinician, n=42). Psychosocial measurements and institutional behavioral records were used to measure outcomes at a 12-week post-intervention point and six months of follow-up to evaluate the effect of the intervention. There were repeated measures ANOVA, regression, and mediation analyses. Social communication (F (2,162) = 8.94, p < 0.001) and emotional regulation (F (2,162) = 7.63, p <0.01) had significant time x group interaction effects. Big effect sizes in the post-test (Cohen’s d = 1.21 and d = 1.08, respectively) and the gains at follow-up were found. Incidents of behavioral engagement were a significant predictor of behavioral incident reductions (β = −0.42, p < 0.01). The peer-led group showed a 56% decrease in behavioral episodes as compared to 15% in the control group, and continued to improve after six months. Collaborative learning led by peers had a considerable positive impact on socio-emotional functionality and minimized the behavioral risk of neurodivergent forensic service users. Their results suggest the incorporation of neurodiversity-informed, peer-mediated methodology into the context of forensic rehabilitation to promote responsivity, engagement, and desistance in the long run.

The study conducted a comprehensive analysis of the production costs of broiler chicken meat and the live performance metrics of broiler flocks across a sample of 100 poultry enterprises located in the Surkhandarya region during the period of September to October 2025. The study estimated the average cost of production per kilogram live weight of broiler to be 35,690 UZS. The analysis revealed that feed costs and chick procurement costs constituted the most significant components of the total production cost per kilogram of live broiler weight, accounting for 69.2% and 19.6%, respectively. The results indicated that the cost of producing one kilogram of live broiler weight in the Surkhandarya region is approximately 11% to 13% higher compared to other regions of Uzbekistan. Keywords: Broiler Meat Production, Poultry Farming, Poultry Enterprises, Food Security, Production Cost Analysis, Economic Performance

Zero-day intrusion detection is still a difficult task because of the difference between high laboratory precision and real-time deployability under strict operational constraints. This paper proposes a lightweight two-stage cascade architecture that is specifically designed for CPU-only environments and strict zero-day evaluation. The proposed architecture only uses statistical and flow-level metadata attributes, which are independent of payload analysis, to ensure compatibility with encrypted traffic. The first stage of the proposed architecture is precision oriented to detect potentially malicious traffic with a low decision threshold, and the second stage is precision oriented to enhance classification and remove false positives. To avoid optimistic bias, a strict attack-type separation protocol is employed, where testing attack types are strictly prohibited from training. The proposed method is tested on three benchmark datasets: CSIC 2012 (HTTP level), UNSW-NB15 (intra-domain), and CSE-CIC-IDS2018 (cross-domain). The experimental results show the excellent intra-domain zero-day detection capability (up to 94.81% accuracy with 0.50% FPR), controllable performance degradation in the cross-domain setting (80.53% accuracy with near-zero FPR), and extremely low FP rates on all datasets. The system provides microsecond-level inference latency (0.002–0.006 ms), a throughput of up to 470,000 requests per second, and memory usage below 6.2 MB without GPU support. These results confirm the significance of architectural optimization and thorough evaluation in building efficient zero-day detection systems.