Development of NIRS prediction models for fineness analysis of vicuna and llama textile fibers

Abstract

The production of llama and vicuna textile fibres is one of the more valuable environmental, cultural and economic activity in the High-Andean region of Catamarca. It is necessary to develop a technology that allows to analyze this material quickly and sustainably. In this work, different treatments and NIRS prediction models were evaluated through R2, SECV, SEV and RPD statistics to predict Fineness in samples of llama and vicuna. The results showed that the best models were obtained with none or minimum treatment. Trough R2 <0.7 and RPD≅1.7 results, we concluded that it is necessary to increase the sample size and explore the development of alternative regression techniques.