Journal of Information Technology and Sciences

Bagging-Based Ensemble of Optimized SVM Classifiers for Robust Breast Cancer Prediction

Satish Kumar Kalagotla — 2026-04-10

Background: Single Support Vector Machine (SVM) classifiers, even when optimized for feature selection and parameters, suffer from high variance and sensitivity to variations in the training data, limiting their reliability in critical medical diagnosis applications. Ensemble methods, particularly bagging, offer a powerful approach to improve robustness and accuracy by combining multiple diverse classifiers.

Objective: This paper proposes a novel heterogeneous bagging ensemble framework that integrates five optimized SVM variants—DT-SVM (missing value handling), Correlation-SVM (multicollinearity-aware), ABC-SVM (feature-optimized), GS-GA-SVM (parameter-optimized), and Standard SVM—to achieve robust and accurate breast cancer prediction.

Methods: The proposed framework employs bootstrap sampling to generate diverse training sets for each base learner. Each SVM variant is trained on bootstrap samples with out-of-bag validation, and predictions are aggregated via weighted voting, with weights optimized using validation performance. The framework was evaluated on four benchmark medical datasets (Wisconsin Breast Cancer, PIMA Indian Diabetes, Hepatitis, and Mammographic Mass) and compared against individual base learners and homogeneous bagging ensembles using 10-fold cross-validation with five repeats.

Results: The heterogeneous bagging ensemble achieved 98.76% accuracy on the Wisconsin dataset, significantly outperforming individual SVM variants (average 95.8%) and standard bagging with homogeneous SVMs (97.1%). The ensemble reduced prediction variance by 67.7% compared to single classifiers (standard deviation 0.0042 vs 0.013). Diversity analysis revealed a moderate correlation among base learners (mean Q-statistic of 0.52 and a mean correlation of 0.65), confirming complementary strengths—optimized weighting assigned the highest weights to ABC-SVM (0.24) and GS-GA-SVM (0.23). Cross-dataset validation showed consistent improvements: PIMA Indian Diabetes (88.67%), Hepatitis (89.51%), and Mammographic Mass (90.83%). Robustness testing demonstrated superior performance under label noise, with only 5.9% degradation at 20% noise compared to 10.0% for standard SVM.

Conclusion: The heterogeneous bagging ensemble of optimized SVMs provides a robust, high-performance framework for breast cancer prediction, significantly reducing variance while improving accuracy. The diversity among base learners and optimized weighting scheme contribute to superior generalization, making it suitable for clinical deployment where prediction stability is paramount.

Designing the Commerce Stack for Autonomous Agents: Transforming Payments, Risk, and Identity

Arjun Wadwalkar — 2026-03-11

Autonomous software agents are increasingly executing commercial decisions that were historically authorized, interpreted, and governed by humans. Existing commerce stacks, including payment rails, fraud systems, identity frameworks, and dispute processes, are structurally optimized for human intent and manual oversight. As a result, they fail to support high-frequency, delegated, and machine-executed transactions without introducing unacceptable risk, ambiguity in liability, and governance friction. This paper argues that autonomous commerce requires a distinct, execution-oriented commerce stack that treats authorization, risk, identity, liability, and governance as machine-enforceable primitives rather than external controls. The paper proposes the Autonomous Commerce Stack Framework (ACSF), a layered architecture designed to support agent-initiated transactions through policy-bound execution, continuous behavioral risk monitoring, machine-native identity, and embedded accountability mechanisms. Unlike prior work that treats payments, risk, and identity in isolation, ACSF integrates these concerns into a unified system in which governance and liability are enforced at transaction time rather than retroactively. The framework is developed through a systematic synthesis of interdisciplinary literature spanning payment systems, autonomous agents, digital identity, and AI governance, and refined through semi-structured interviews with domain experts in payments, risk, and enterprise automation. Findings highlight fundamental breakdowns in human-centric authorization models, static identity systems, and post hoc liability regimes when applied to autonomous agents. The ACSF offers a concrete architectural blueprint and transaction semantics to guide the design of scalable, auditable, and policy-compliant autonomous commerce systems.

IoT and Machine Learning–based Maternal Health Monitoring System

Dhanya Priya S. L — 2026-01-19

Monitoring a baby’s health during pregnancy matters not just for the baby but for the mother as well. Doctors rely on tools like cardiotocography (CTG) and fetal electrocardiography (fECG) to spot trouble, whether it’s distress, hypoxia, or developmental issues. But these old-school methods trip up when it comes to interpreting signals, dealing with noise, or just getting access when and where they are needed. Now, the landscape is shifting. Artificial intelligence and machine learning are stepping in, along with smarter sensors and electronics. Suddenly, it’s possible to monitor fetal health more accurately, around the clock, and without being invasive. AI isn’t just a buzzword here it sharpens CTG analysis, lets us read signals with more confidence, and shrugs off a lot of the noise that used to muddy the waters. Deep learning and fuzzy logic help make sense of messy data. On top of that, new tech like fibre Bragg grating sensors, adaptive noise cancellation, and secure telemonitoring are changing the game. Healthcare providers can now keep an eye on both mother and baby in real time even when they’re not in the clinic thanks to IoT-connected wearables. Digital twin models and the use of AI in genomic analysis crank things up another notch. They open the door to predictive and personalized care, tailored to the specific needs of each mother and child. Of course, there’s still work to do cleaning up data, making sure results are understandable, and proving these tools actually work in the clinic. But as AI-driven algorithms, secure IoT systems, and advanced sensors come together, prenatal care is moving toward a future that’s more personal, more accurate, and a lot more accessible.

Blockchain Technology in Digital Advertising: Opportunities and Challenges

Devarakonda Kanaka Mahalakshmi Devi — 2026-03-28

Blockchain technology is transforming the digital advertising industry by improving transparency, security, and efficiency across the advertising ecosystem. Traditional digital advertising systems often suffer from issues such as ad fraud, lack of transparency, and misuse of user data, which reduce trust among advertisers, publishers, and consumers. Blockchain introduces a decentralized and immutable ledger that records advertising transactions, including ad impressions, clicks, and payments, ensuring that the data cannot be altered or manipulated. This helps advertisers verify genuine engagement and significantly reduces fraudulent activities such as fake clicks and bot traffic. Smart contracts further enhance the system by automating advertising agreements and releasing payments only when predefined conditions are met, ensuring fair compensation for publishers. In addition, blockchain empowers users by giving them greater control over their personal data and enabling consent-based data sharing. Token-based systems can also reward users for interacting with advertisements, creating a more balanced and user-centric advertising environment. Despite these benefits, several challenges still limit large-scale adoption, including scalability issues, high computational costs, integration difficulties with existing advertising infrastructure, and regulatory uncertainties. Nevertheless, with continuous technological advancements and integration with emerging technologies such as artificial intelligence, Web3, and decentralized autonomous organizations, blockchain has the potential to create a more transparent, secure, and trustworthy digital advertising ecosystem.

Strategic Integration of Inclusive Utilities into Digital Public Infrastructure (DPI)

Tanu — 2026-01-31

Digital Public Infrastructure (DPI) has emerged as a foundational layer for delivering population-scale digital services. However, most existing systems emphasize technical efficiency while marginalizing social inclusion and accessibility. This paper evaluates the integration protocols required to transition socio-technical algorithms from theoretical models to functional public service utilities. Focusing on Integration, the research delineates a multi-tiered architecture designed to bridge the gap between ICT metrics and social equity. By utilizing a synchronized data persistence layer, the framework ensures that inclusive rankings are not merely transient calculations but become a permanent, auditable part of the Digital Public Infrastructure. The results demonstrate that a structured integration approach allows for real-time policy monitoring and enhances the transparency of digital public services. The study is structured around three core objectives: first, minimizing the digital divide by mathematically rewarding content accessibility; second, envisaging and delivering digital utilities that adapt to varying user infrastructures and experience levels, and third, engineering the technological protocols required to integrate these utilities into a persistent and auditable Digital Public Infrastructure.