http://localhost:8080/xmlui/handle/123456789/37 Wed, 09 Sep 2026 23:51:34 GMT 2026-09-09T23:51:34Z http://localhost:8080/xmlui/handle/123456789/2421 EVAPOTRANSPIRATION ESTIMATION USING EMPIRICAL AND MACHINE LEARNING MODELS A COMPARATIVE ASSESSMENT OF ACCURACY AND TRANSFERABILITY AFNA HAMEEM H; ABHIRAM M A; SREENSNDANA P V; Dr. ASHA JOSEPH, (Guide) This study compared empirical and machine learning (ML) models for estimating reference evapotranspiration (ET₀) at two agroclimatically contrasting stations in Kerala: RARS Pattambi (inland, semi-arid) and the College of Agriculture, Vellayani (coastal, humid). Using thirty years (1995–2025) of daily meteorological data, the FAO-56 Penman–Monteith equation served as the benchmark. Five ML models - CNN, LSTM, CNN-LSTM,XGBoost, and Random Forest—were developed under four input configurations (full variable, temperature-based, mass transfer-based, and radiation-based) and compared against six empirical models, with accuracy assessed via R², RMSE, and MAE, and spatial transferability quantified through a transferability score. ML models consistently outperformed empirical approaches across all input categories.With complete inputs, all ML models achieved R² > 0.95, with XGBoost RF performing best. Under reduced inputs, radiation-based models performed strongest, with LSTM and XGBoost exceeding an R² of 0.87, while the empirical Makkink model also performed comparatively well, reflecting the strong physical link between solar radiation and ET₀ in Kerala's climate. Temperature-based models performed weakest overall. Transferability analysis showed that models trained at Pattambi transferred more effectively to Vellayani than the reverse, particularly for radiation-based inputs, where external validation exceeded internal accuracy. Mass transfer-based models exhibited the weakest cross-station transferability, reflecting differing humidity dynamics between the inland and coastal environments. Validation using 2025 data confirmed these patterns. Overall, ML models - particularly XGBoost and Random Forest with complete data, and LSTM/CNN-LSTM under radiation-based inputs - offer a reliable and transferable framework for ET₀ estimation. Radiation-based inputs are recommended for data-scarce settings, and the demonstrated inland-to-coastal transferability supports extending these models to ungauged locations, providing a scientific foundation for improved irrigation planning and water resource management across Kerala. Tue, 08 Sep 2026 00:00:00 GMT http://localhost:8080/xmlui/handle/123456789/2421 2026-09-08T00:00:00Z http://localhost:8080/xmlui/handle/123456789/2420 Development of a Learning Based Shele- Life Prediction Model for Kaju Katli Aleena Biju; Sourav P; Navya Ankita Anand; Abhishek, (Guide) S Kaju katli is a widely consumed traditional Indian confectionery prepared primarily from cashew nuts and sugar. Despite its popularity, the product has a limited shelf life due to physicochemical and microbiological deterioration during storage, including moisture loss, lipid oxidation, and microbial growth. Conventional shelf-life determination methods rely on repeated laboratory testing, which is time-consuming, labour-intensive, and expensive. This study aims to develop a machine learning-based approach for predicting the shelf life of kaju katli using quality parameters obtained through laboratory analysis. The work focuses on integrating food science with artificial intelligence to provide a rapid, reliable, and cost-effective alternative to conventional shelf-life estimation. The study evaluates key physicochemical and microbiological parameters, namely moisture content, water activity (aw), free fatty acid (FFA), standard plate count (SPC), and yeast and mould count (YMC), as indicators of product quality during storage. Experimental data collected from laboratory analyses are organized and enhanced through feature engineering techniques to improve dataset quality and predictive capability. A Random Forest regression model is developed to establish relationships between these quality attributes and the remaining shelf life of kaju katli. The model is trained and evaluated using standard performance metrics to ensure prediction accuracy and reliability. The proposed machine learning model is expected to accurately predict shelf life by capturing the complex interactions among physicochemical and microbiological parameters that influence product deterioration. The approach minimizes dependence on prolonged storage studies while providing a scientific basis for quality monitoring and decision-making. The findings are expected to demonstrate that moisture content, water activity, FFA, SPC, and YMC are significant predictors of shelf life and can effectively support intelligent food quality assessment. Overall, this research highlights the potential of machine learning in modern food engineering by offering an efficient tool for shelf-life prediction of traditional confectionery products. The developed model can help manufacturers optimize storage conditions, reduce food wastage, improve inventory management, and enhance consumer confidence by providing more accurate shelf-life estimates. Furthermore, the methodology can be extended to other traditional Indian sweets and similar food products, contributing to the advancement of intelligent food processing and quality assurance systems. Tue, 08 Sep 2026 00:00:00 GMT http://localhost:8080/xmlui/handle/123456789/2420 2026-09-08T00:00:00Z http://localhost:8080/xmlui/handle/123456789/2418 ULTRASOUND ASSISTED ACCELERATION OF WINE AGING Souvmia Swaminathan; Evin T S; Adish T; Khaleel Mohammed Hasan V; ABHISHEK.S, (Guide) Thu, 01 Jan 2026 00:00:00 GMT http://localhost:8080/xmlui/handle/123456789/2418 2026-01-01T00:00:00Z http://localhost:8080/xmlui/handle/123456789/2408 DEVELOPMENT OF TURMERIC LEAF ESSENTIAL OIL INCORPORATED ACTIVE PACKAGING FILM FOR FOOD APPLICATIONS ARSHAD K; ASWATHI S; MANASA C; VITHU PRABHA, (Guide) The present study aimed to develop and evaluate a biodegradable active packaging film by incorporating turmeric leaf essential oil (TLEO) into a starch-chitosan matrix for food packaging applications. Turmeric leaves, an underutilized agricultural by-product, were utilized as a sustainable source of bioactive essential oil. TLEO was extracted by hydrodistillation using a Clevenger apparatus and incorporated into starch-chitosan films at concentrations of 0, 1, 2, 3, and 4% (v/v) using the solvent-casting method. The developed films were characterized for mechanical and physical properties, including thickness, force at break, elongation at break, bursting strength, and colour. Their preservation efficacy was evaluated by determining the total plate count (TPC) of refrigerated sardine samples wrapped with the films, while insect-repellent activity against Sitophilus oryzae was also assessed. The incorporation of TLEO significantly improved the mechanical performance of the films. Film thickness increased from 0.084 to 0.208 mm, bursting strength from 23.67 to 30.33 psi, force at break from 1.37 to 5.83 N, and elongation at break from 0.90 to 9.00 mm as TLEO concentration increased to 4%. Colour analysis showed a gradual decrease in lightness and an increase in yellowness with increasing essential oil concentration, confirming the successful incorporation of TLEO into the polymer matrix. Antimicrobial evaluation revealed a concentration-dependent reduction in microbial growth during refrigerated storage of sardine. The film containing 4% TLEO exhibited the best preservation performance, maintaining the total plate count at 6.73 log CFU g⁻¹ after seven days of storage compared with 7.51 log CFU g⁻¹ for the control film. Repellency studies also demonstrated enhanced insect-repellent activity with increasing TLEO concentration. Based on the combined evaluation of mechanical properties, antimicrobial efficacy, and insectrepellent activity, the starch-chitosan film containing 4% TLEO was identified as the optimum formulation. The study demonstrates that turmeric leaf essential oil, obtained from agricultural waste, can be effectively utilized to produce environmentally friendly active packaging films with improved mechanical strength, antimicrobial activity, and insect-repellent properties, highlighting their potential as sustainable alternatives to conventional petroleum-based food packaging materials. Thu, 01 Jan 2026 00:00:00 GMT http://localhost:8080/xmlui/handle/123456789/2408 2026-01-01T00:00:00Z