Data Mining for Prediction of Clothing Insulation

Owing to difficulties of gathering large volumes of textile domain data in a context of less mining research, predicting the characteristics of garments becomes an important open problem which receives more and more attention from the textiles research community. In this research work, the field of Data mining attempts to predict clothing insulation factors with the goal of understanding the computational character of learning. Characteristics of clothing learning is being investigated as a technique for making the selection and usage of training data and their outcomes. It is observed from the results obtained by experimentation that the Linear Regression is quiet appealing because of effectiveness in terms of high prediction rate and Linear Regression is able to discover the clothing insulation performance in a most efficient manner in comparison to all other leaning algorithms experimented. Data mining Classifiers has showed spectacular success in reducing classification error from learned classifiers like Linear regression, LeastMedSq and AdditiveRegression functions have been analyzed for improving the predictive power of classifier learning systems.