Deep learning (DL) enables the creation of computational models comprising multiple processing layers that learn data representations at multiple levels of abstraction. In the recent past, the use of deep learning has been proliferating, yielding promising results in applications across a growing number of fields, most notably in image processing, medical image analysis, data analysis, and bioinformatics. DL algorithms have also had a significant positive impact through yielding improvements in screening, recognition, segmentation, prediction, and classification applications across different domains of healthcare, such as those concerning the abdomen, cardiac, pathology, and retina. Given the extensive body of recent scientific contributions in this discipline, a comprehensive review of deep learning developments in the domain of diabetic retinopathy (DR) analysis, viz., screening, segmentation, prediction, classification, and validation, is presented here. A critical analysis of the relevant reported techniques is carried out, and the associated advantages and limitations highlighted, culminating in the identification of research gaps and future challenges that help to inform the research community to develop more efficient, robust, and accurate DL models for the various challenges in the monitoring and diagnosis of DR.

In the field of computational linguistics, addressing machine translation (MT) challenges for low-resource languages remains crucial, as these languages often lack extensive data compared to high-resource languages. General large language models (LLMs), such as GPT-4 and Llama, primarily trained on monolingual corpora, face significant challenges in translating low-resource languages, often resulting in subpar translation quality. This study introduces Language-Specific Fine-Tuning with Low-rank adaptation (LSFTL), a method that enhances translation for low-resource languages by optimizing the multi-head attention and feed-forward networks of Transformer layers through low-rank matrix adaptation. LSFTL preserves the majority of the model parameters while selectively fine-tuning key components, thereby maintaining stability and enhancing translation quality. Experiments on non-English centered low-resource Asian languages demonstrated that LSFTL improved COMET scores by 1-3 points compared to specialized multilingual machine translation models.Additionally, LSFTL’s parameter-efficient approach allows smaller models to achieve performance comparable to their larger counterparts, highlighting its significance in making machine translation systems more accessible and effective for low-resource languages.

In typical e-commerce warehouse operations, upon receiving the orders from customers, the purchased items need to be retrieved from shelves and then packaged accordingly for delivery. To automate and speed up the item retrieval process, a Smart Warehouse usually employs a management system, called the Automated Retrieval System (ARS), to control and schedule the retrieval jobs. The working principle of ARS is crucial to the Smart Warehouse because it will have a great impact on the subsequent downstream processes. In short, all the items in a particular order should be considered as an integral part; if one of these items encounters a much larger retrieval delay than others do, then the entire order may experience an unnecessary latency. In the past, the integrality of order has not received much attention for the parallel retrieval process of multiple stackers. To take this into account, this paper proposes using an Order Tag to label all the items that belong to the same order for retrieval job scheduling. The way of calculating the Order Tags will then determine the scheduling discipline of the ARS. With the objectives of minimizing the average delay and ensuring the fairness, two algorithms are proposed. They are named as Dynamic Order-Based (DOB) and Dynamic Order-Based with Threshold (DOBT) Scheduling Algorithms, respectively. Compared with the First-Come-First-Serve and other approaches, the simulation results show that DOB and DOBT are able to reduce the average order retrieval delay by at least 30%, and generate less backlog pressure to the downstream operations.

One of the key challenges in financial distress data is class imbalance, where the data are characterized by a highly imbalanced ratio between the number of distressed and non-distressed samples. This study examines eight resampling techniques for improving distress prediction using the XGBoost algorithm. The study was performed on a dataset acquired from the CSMAR database, containing 26,383 firm-quarter samples from 639 Chinese A-share listed companies (2007–2024), with only 12.1% of the cases being distressed. Results show that standard Synthetic Minority Oversampling Technique (SMOTE) enhanced F1-score (up to 0.73) and Matthews Correlation Coefficient (MCC, up to 0.70), while SMOTE-Tomek and Borderline-SMOTE further boosted recall, slightly sacrificing precision. These oversampling and hybrid methods also maintained reasonable computational efficiency. However, Random Undersampling (RUS), though yielding high recall (0.85), suffered from low precision (0.46) and weaker generalization, but was the fastest method. Among all techniques, Bagging-SMOTE achieved balanced performance (AUC 0.96, F1 0.72, PR-AUC 0.80, MCC 0.68) using a minority-to-majority ratio of 0.15, demonstrating that ensemble-based resampling can improve robustness with minimal impact on the original class distribution, albeit with higher computational cost. The compared findings highlight that no single approach fits all use cases, and technique selection should align with specific goals. Techniques favoring recall (e.g., Bagging-SMOTE, SMOTE-Tomek) are suited for early warning, while conservative techniques (e.g., Tomek Links) help reduce false positives in risk-sensitive applications, and efficient methods such as RUS are preferable when computational speed is a priority.

In all-optical packet switching, packets may arrive at an optical switch in an uncoordinated fashion. To prevent packet loss in the switch, fiber delay lines (FDLs) are used as optical buffers to store optical packets. However, assigning FDLs to the arrival packets to achieve high throughput, low delay, and low loss rate is not a trivial task. In the authors' companion paper, several efficient scheduling algorithms were proposed for single-stage shared-FDL optical packet switches (OPSs). To further enhance the switch's scalability, this work was extended to a multistage case. In this paper, two scheduling algorithms are proposed: 1) sequential FDL assignment and 2) multicell FDL assignment algorithms for a three-stage optical Clos-Network switch (OCNS). The paper shows by simulation that a three-stage OCNS with these FDL assignment algorithms can achieve satisfactory performance.