Faculty of Computing

Permanent URI for this communityhttps://rda.sliit.lk/handle/123456789/4038

Browse

Filters

20252

Advanced Search

Filter by

Settings

Start date: 2025

Search Results

Now showing 1 - 2 of 2
  • Thumbnail Image
    PublicationOpen Access
    Exploring nontoxic perovskite materials for perovskite solar cells using machine learning
    (Discover, 2025-07-06) Pabasara, W.G.A.; Wijerathne, H.A.H.M; Karunarathne, M.G.M.M.; Sandaru, D.M.C.; Abeygunawardhana, Pradeep K. W.
    erovskite solar cells are promising renewable energy technology that faces significant challenges due to the Pb induced toxicity. The current study addresses this issue by leveraging machine learning techniques to explore Pb-free perovskite materials that ensure environmental sustainability and human safety. A highly accurate machine learning model was developed to predict Goldschmidt factor and the band gap, aiming to discover lead-free perovskites. Extreme Gradient Boost (XGBoost), Random Forest (RF), Gradient Boost Regression (GBR), and Ada Boost Regression (ABR) models were employed for this purpose. The findings exhibit that XGBoost delivers the most precise and reliable results for Goldsmith tolerance factor prediction with an accuracy of 98.5%. Furthermore, GBR model, combined with K-nearest neighbors (KNN) model delivers an impressive accuracy of 98.7% for the band gap predictions. 49 Pb-free perovskite materials were screened out considering the toxicity and the abundance. Utilizing Principal Component Analysis (PCA) and K-means clustering, six optimal materials (KBiBr3, KZnBr3, RbBiBr 3, RbZnBr3, MAGeI3, and FAGeI3null) were identified as the potential environment-friendly materials for photovoltaic applications. These results show the crucial role of machine learning and statistical analysis in discovering nontoxic and environmental-friendly perovskite materials, advancing the development of sustainable energy solutions.
  • Thumbnail Image
    PublicationOpen Access
    A novel application with explainable machine learning (SHAP and LIME) to predict soil N, P, and K nutrient content in cabbage cultivation
    (Elsevier B.V., 2025-08) Abekoon, T; Sajindra, H; Rathnayake, N; Ekanayake, I.U.; Jayakody, A; Rathnayake, U
    Cabbage (Brassica oleracea var. capitata) is commonly cultivated in high altitudes and features dense, tightly packed leaves. The Green Coronet variety is well-known for its robust growth and culinary versatility. Maximizing yield is crucial for food sustainability. It is essential to predict the soil's major nutrients (nitrogen, phosphorus, and potassium) to maximize the yield. Artificial intelligence is widely used for non-linear predictions with explainability. This research assessed the predictive capabilities of soil nitrogen, phosphorus, and potassium levels with explainable machine learning methods over an 85-day cabbage growth period. Experiments were conducted on cabbage plants grown in central hills of Sri Lanka. SHapley Additive exPlanations (SHAP) and Local Interpretable Model-agnostic Explanations (LIME) were used to clarify the model's predictions. SHAP analysis showed that high feature values of the number of days and plant average leaf area negatively impacted for nutrient predictions, while high feature values of leaf count and plant height had a positive effect on the nutrient predictions. To validate the results, 15 greenhouse-grown cabbage plants at various growth stages were selected. The nitrogen, phosphorus, and potassium levels were measured and compared with the predicted values. These insights help refine predictive models and optimize agricultural practices. A user-friendly application was developed to improve the accessibility and interpretation of predictions. This tool is a user-friendly platform for end-users, enabling effective use of the model's predictive capabilities.