Time Series Forecasting using Vector quantization

Abstract

Time series data prediction comprises of linear and non linear prediction techniques, where the non-linear time series prediction techniques will drastically enhance when compared with performance to linear time series prediction models. In this paper I present improvised complexity and processing time acquired that often disallows proficient utilization of nonlinear tools by emphasizing more on basic nonlinear procedures for performing time series forecasting using vector quantization techniques for predicting the data value which is considered to be the missing data [1]. The methods implemented towards vector quantization are represented to be congruent with similar missing data as the paper provides alternate methods that are more complex than prediction tools for handling the data using tolerable complexity with acceptable processing time.