华美银行

Sequential recommender systems must model long-range user behavior while operating under strict memory and latency constraints. Transformer-based approaches achieve strong accuracy but suffer from quadratic attention complexity, forcing aggressive truncation of user histories and limiting their practicality for long-horizon modeling. This paper presents HoloMambaRec, a lightweight sequential recommendation architecture that combines holographic reduced representations for attribute-aware embedding with a selective state space encoder for linear-time sequence processing. Item and attribute information are bound using circular convolution, preserving embedding dimensionality while encoding structured metadata. A shallow selective state space backbone, inspired by recent Mamba-style models, enables efficient training and constant-time recurrent inference. Experiments on Amazon Beauty and MovieLens-1M datasets demonstrate that HoloMambaRec consistently outperforms SASRec and achieves competitive performance with GRU4Rec under a constrained 10-epoch training budget, while maintaining substantially lower memory complexity. The design further incorporates forward-compatible mechanisms for temporal bundling and inference-time compression, positioning HoloMambaRec as a practical and extensible alternative for scalable, metadata-aware sequential recommendation.

Risk-aware portfolio optimization requires balancing return maximization with strict control of tail risk, drawdowns, and regulatory exposure constraints under non-stationary market conditions. While deep reinforcement learning has shown promise for portfolio management, existing approaches typically optimize expected returns or incorporate risk preferences through soft reward shaping, offering no guarantees on constraint satisfaction. This paper presents a constrained distributional policy gradient framework for portfolio optimization that explicitly integrates distributional return modeling with hard risk constraints. The proposed method learns the full return distribution using implicit quantile networks and enforces constraints on Conditional Value-at-Risk (CVaR), maximum drawdown, and asset exposure limits through Lagrangian-based policy gradient optimization. Risk constraints are evaluated directly over the learned return distribution, enabling differentiable and distribution-aware constraint enforcement without reliance on sample-based estimates. Empirical evaluation on real market data from 2010-2023 demonstrates that the proposed approach achieves competitive cumulative returns, improved risk-adjusted performance, and lower constraint violation rates compared to unconstrained and reward-shaped reinforcement learning baselines. The learned return distributions adapt dynamically to changing market regimes, providing interpretable risk quantification and improved robustness during periods of market stress, including the COVID-19 crash. These results highlight the effectiveness of constrained distributional policy gradient methods as a practical foundation for risk-aware portfolio optimization.

Enterprises increasingly adopt multi cloud architectures to take advantage of diverse database engines, regional availability, and cost models. In these environments, ETL pipelines must process large, distributed datasets while minimizing latency and transfer cost. Push down optimization, which executes transformation logic within database engines rather than within the ETL tool, has proven highly effective in single cloud systems. However, when applied across multiple clouds, it faces challenges related to data movement, heterogeneous SQL engines, orchestration complexity, and fragmented security controls. This paper examines the feasibility of push down optimization in multi cloud ETL pipelines and analyzes its benefits and limitations. It evaluates localized push down, hybrid models, and data federation techniques that reduce cross cloud traffic while improving performance. A case study across Redshift and BigQuery demonstrates measurable gains, including lower end to end runtime, reduced transfer volume, and improved cost efficiency. The study highlights practical strategies that organizations can adopt to improve ETL scalability and reliability in distributed cloud environments.

Kubernetes provides native autoscaling mechanisms, including the Horizontal Pod Autoscaler, Vertical Pod Autoscaler, and node-level autoscalers, to enable elastic resource management for cloud-native applications. However, production environments frequently experience Service Level Objective violations and cost inefficiencies due to reactive scaling behavior, limited use of application-level signals, and opaque control logic. This paper investigates how Kubernetes autoscaling can be enhanced using AIOps principles to jointly satisfy SLO and cost constraints under diverse workload patterns without compromising safety or operational transparency. We present a gap-driven analysis of existing autoscaling approaches and propose a safe and explainable multi-signal autoscaling framework that integrates SLO-aware and cost-conscious control with lightweight demand forecasting. Experimental evaluation using representative microservice and event-driven workloads shows that the proposed approach reduces SLO violation duration by up to 31 percent, improves scaling response time by 24 percent, and lowers infrastructure cost by 18 percent compared to default and tuned Kubernetes autoscaling baselines, while maintaining stable and auditable control behavior. These results demonstrate that AIOps-driven, SLO-first autoscaling can significantly improve the reliability, efficiency, and operational trustworthiness of Kubernetes-based cloud platforms.

The exponential growth of heterogeneous data sources across modern cloud ecosystems has heightened the need for scalable, governed, and cost-efficient data integration architectures. This paper presents a comprehensive implementation of the Medallion Architecture on Amazon Web Services (AWS) using dbt (Data Build Tool) and Apache Iceberg to deliver an ACID-compliant, modular, and production-grade data lakehouse. The proposed framework integrates Redshift, Athena, Glue, and Managed Workflows for Apache Airflow (MWAA) to support structured ELT workflows across bronze, silver, and gold layers, enabling robust data lineage, automated testing, and seamless schema evolution. A large-scale retail analytics case study demonstrates substantial operational benefits, including 60–75% faster development cycles, up to 50% query cost reduction, and significantly improved data quality compared with traditional Glue-based ETL pipelines. By combining SQL-centric modeling with Iceberg’s transactional capabilities, this work provides quantitative evidence, best practices, and design patterns for building scalable, secure, and future-ready cloud data platforms. The resulting architecture serves as a practical blueprint for organizations seeking to modernize legacy pipelines and adopt governed, high-performance lakehouse ecosystems on AWS.