Association of Indonesian Tin Exporters (AETI) – Tin price forecasting workshop
This course was carried out from 20 to 22 November 2022 in Safran Hotel, Pangkal Pinang, Bangka Belitung. This course discussed the long-term forecasting methods for tin prices. This course contained the statistical and econometrical analysis, econometric approach, time-series approach and stochastic approach to forecast future tin prices. This course also discussed the forecasting performance measures using RMSE, MAE and MAPE.
Indonesian Society of Economic Geologists (MGEI)
This course discussed from the conventional to the most sophisticated models to forecast future energy & metal prices. The participants come from Bumi Resources Mining, Toba Sejahtera, Medco, Petrosea and Banpu. The econometric model was introduced to forecast future prices based on the correlation and cointegration to the relevant variables. This correlation was examined using the PCC and the spearman rank and the cointegrations used the vector error correction (VEC) model. In this case, the econometric model to forecast future prices used the multiple linear regression (MLR) model and the weighted least square (WLS) model.
The time-series was carried out to determine the potential time-series repetition in the future. This proposed model for the time-series was the autoregressive integrated moving average (ARIMA) that has been recognized as one of the best time-series models to forecast future prices. This course also introduced the metaheuristic model (e.g., bat algorithm and grey wolf algorithm) to optimize the volatility. The machine learning system was introduced such as the artificial neural network, multilayer perceptron and SHAP models. Finally, these models were tested to determine the most accurate model to forecast future coal prices.
ICI Panel Gathering
This presentation was held at Westin Hotel Jakarta. It investigated the correlation between the Indonesian Coal Index (ICI) to the South African and Australian coal prices. This presentation observed the statistical and the econometrical analysis to detect the data distribution, stationary and its cointegration to another relevant variable between the Indonesian Coal Index (ICI) to oil, gold, copper, global economic & policy uncertainty (GEPU) index and geopolitical risk (GPR) index. Based on these investigations, the forecasting models such as the geometric Brownian motion (GBM) model and the artificial neural network (ANN) were selected. Therefore, this can conclude future coal prices using these proposed models.
INDONESIA MINER
This workshop was held at Westin Hotel Jakarta from 4 – 5 June 2024. This workshop discussed the latest methods in forecasting energy and metal prices. the stochastic models were introduced to accept the noisy, random and non-linear data. The artificial intelligence was discussed and its application in metal and energy forecasting. To optimize the artificial intelligence model, the hybrid models were taken into account. This combined the artificial neural network (ANN) and the bio-inspired algorithm or swarm intelligence (SI) models. The SI models discussed in this workshop were the bat algorithm (BA), whale optimization algorithm (WOA) and grey wolf optimizer (GWO).
