Journal Articles

Permanent URI for this collectionhttps://mro.massey.ac.nz/handle/10179/7915

Browse

Filters

20252

Settings

Author: search.filters.author.Ding LHas files: Yes

Search Results

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Cooling demand reduction with nighttime natural ventilation to cool internal thermal mass under harmonic design-day weather conditions
    (Elsevier Ltd, 2025-02-01) Li M; Shen X; Wu W; Cetin K; Mcintyre F; Wang L; Ding L; Bishop D; Bellamy L; Liu M
    Cooling demand is steadily increasing across different climate zones due to global warming. A potential solution for cooling demand reduction is applying nighttime natural ventilation to cool internal thermal mass. However, a simplified and accurate modelling framework to assess the technique is still missing. The goal of the study is to build that framework integrated with a validated internal thermal mass model and apply the framework to quantify the cooling demand reduction potential in a space with different thermal mass and envelope configurations and in different climate zones. Results show that using Granite as internal thermal mass is three times more effective than concrete to reduce peak cooling load. Adding too much internal thermal mass can create adverse effects on cooling load reduction. The optimum thickness of internal thermal mass is between 28 and 45 mm. Envelope construction also has an influence on the performance of nighttime cooling. Applying the technique in buildings with lightweight structures reduces peak cooling load by 35.9% more than heavyweight structures. As heavyweight structures delay the release of the daily absorbed heat and cause higher indoor air temperatures at night. The two belts between the Tropic of Cancer and 60 degrees north latitude, and between the Tropic of Capricorn and 45 degrees south latitude are suitable for nighttime natural ventilation of internal thermal mass, achieving the annual cooling demand reduction above 1.25 kWh m−2. In Dessert climate zones, the technique exhibits an extraordinary potential to reduce cooling demand, up to 6.67 kWh m−2 per year.
  • Thumbnail Image
    Item
    Transformer-based multiple instance learning network with 2D positional encoding for histopathology image classification
    (Springer Nature Switzerland AG, 2025-05) Yang B; Ding L; Li J; Li Y; Qu G; Wang J; Wang Q; Liu B
    Digital medical imaging, particularly pathology images, is essential for cancer diagnosis but faces challenges in direct model training due to its super-resolution nature. Although weakly supervised learning has reduced the need for manual annotations, many multiple instance learning (MIL) methods struggle to effectively capture crucial spatial relationships in histopathological images. Existing methods incorporating positional information often overlook nuanced spatial correlations or use positional encoding strategies that do not fully capture the unique spatial dynamics of pathology images. To address this issue, we propose a new framework named TMIL (Transformer-based Multiple Instance Learning Network with 2D positional encoding), which leverages multiple instance learning for weakly supervised classification of histopathological images. TMIL incorporates a 2D positional encoding module, based on the Transformer, to model positional information and explore correlations between instances. Furthermore, TMIL divides histopathological images into pseudo-bags and trains patch-level feature vectors with deep metric learning to enhance classification performance. Finally, the proposed approach is evaluated on a public colorectal adenoma dataset. The experimental results show that TMIL outperforms existing MIL methods, achieving an AUC of 97.28% and an ACC of 95.19%. These findings suggest that TMIL’s integration of deep metric learning and positional encoding offers a promising approach for improving the efficiency and accuracy of pathology image analysis in cancer diagnosis.