Swarm intelligence algorithms optimized support vector regression with cloud driven framework for predicting jute yarn properties

The physical properties of jute fibres, including root content, defect, bundle strength, and fineness, have a significant influence on yarn properties. In the present study, nine swarm intelligence–based optimization approaches (Cuckoo search algorithm (CSA), Firefly Optimization Algorithm (FFA), Sparrow Search Algorithm (SSA), Harris Hawks Optimizer (HHO), Bat Algorithm (BA), Artificial Bee Colony (ABC), Particle swarm optimization (PSO), Grey wolf optimization (GWO) and Whale optimization (WO)) integrated with Support Vector Regression were employed to estimate jute yarn properties using fibre quality parameters. For model development, a dataset consisting of 414 experimental observations was used, where 70% of the samples were allocated for training and the remaining 30% for testing. The input variables included bundle strength (g/tex), defect (%), root content (%), and fineness (tex), while the output responses targeted for prediction were yarn tenacity (cN/tex) and elongation (%). Among the developed models, WO-SVR model stands out best performing model (For tenacity = R2TR = 0.950 & R2TS = 0.89 and for elongation = R2TR = 0.912 & R2TS = 0.869) both in training and testing phase. Furthermore, WO-SVR had the lowest COM score of 1.228 and demonstrated outstanding predictive accuracy and generalization. An Android App was developed using the WO-SVR model to enable practical implementation and real-time prediction.