Forecasting non-linear WPI of manufacture of chemicals and chemical products in India: an MLP approach

Abstract

Forecasting predicts/ estimates the future by taking into account historical data. It's an instrument of decision-making that assists companies/ enterprises/ institutions in managing uncertainties. It allows the business to set up goals and create a budget. Forecasting helps us to anticipate changes and therefore guide us towards data-driven strategies and reasoned decisions/ choices. It serves as a proactive rather than a reactive measure. There are several forecasting approaches for the univariate time series (TS) data, e.g. *Author for correspondence Auto-Regressive Integrated Moving Average (ARIMA) [1-3], exponential smoothing (ES) [4-6], Support Vector Regression (SVR) [7, 8, 9], neural approaches [10-12]. The recent studies show that the researchers employed these approaches for forecasting various TS data, e.g., stock prices [1, 2, 13], prices of agricultural products [4, 5, 6], price index [10, 14].India's chemical industry is highly diverse, covers a wide range of products, and has a large import and export base. More than two million people are working here. On the economic growth of India, this sector's contribution is notable. In the future, there exists a significant opportunity for the industry's business/ revenue growth.