Banca de DEFESA: MANOEL ALEXANDRE DE LUCENA

Uma banca de DEFESA de MESTRADO foi cadastrada pelo programa.
STUDENT : MANOEL ALEXANDRE DE LUCENA
DATE: 26/07/2024
TIME: 10:00
LOCAL: https://meet.google.com/yji-bear-fuh
TITLE:

Modeling and forecasting small time series of national eco-efficiency

KEY WORDS:

Sustainable development. Environmental and economic impacts. Window Data Envelopment Analysis. Machine learning algorithms.

PAGES: 131
BIG AREA: Ciências Agrárias
AREA: Agronomia
SUMMARY:

Eco-efficiency time series are useful for monitoring the relationship between economic and environmental variables. Thus, forecasting eco-efficiency saves resources and time and provides researchers and managers with insights into future eco-efficiency and the monitoring of environmental policy in different sectors of the economy. However, national time series on eco-efficiency are mostly small or very small. Furthermore, it is relevant to consider models that simultaneously involve all countries, that is, a stacked approach. Thus, the applied pooled approaches can verify whether just a pooled model can predict better than individual time series models for each country. In this context, this research aims to study a method for modeling and forecasting time series of national eco-efficiency. Individual machine learning models for time series are adopted in both cases: Support Vector Regression (SVR), Long Short-Term Memory (LSTM), Decision Tree Regression (DTR); and ensemble: combination by Simple Mean (SA), Simple Median (SM), Minimum Variance (MV); Random Forest Regression (RFR) and Extreme Gradient Boosting (XGB). Furthermore, Considering the individual approach, the Autoregressive Integrated Moving Average (ARIMA) and Exponential Smoothing (ETS) are also considered. In turn, to obtain the national eco-efficiency time series, Data Envelopment Analysis combined with Window Analysis (WDEA) is applied. To calculate the window size in WDEA, a method based on eco-efficiency divergence is proposed. In particular, the ideal window width is the one that maximizes the dispersion of eco-efficiency. The Mercosur, BRICS and G18 countries were considered as case studies, involving annual eco-efficiency time series from 1995 to 2020. In the three cases studied, the The pooled approach won in 50% of the series in Mercosur, 25% in the BRICS and 9.1% in the G18. Particularly, of the 19 best models, 12 (63.1%) were individual models. Furthermore, the average eco-efficiency projected for the next 6 years was low. The results showed that for groups that are possibly more heterogeneous in terms of environmental and economic factors, individual time series analysis wins over the grouped approach. Given the low projected eco-efficiency, agreements and actions adapted to the reality of countries and groups can provide ways to increase the eco-efficiency of countries aligned with sustainable development. Therefore, considering these results, actions can be proposed for policy makers: (i) alignment of goals between countries and groups based on predicted eco-efficiency considering time series models; (ii) global agreement strategies that consider the individual reality of countries in terms of economic and environmental resource endowments; and (iii) use of technology to obtain and use renewable resource sources that reduce greenhouse gas emissions.

COMMITTEE MEMBERS:
Interno - CARLOS WAGNER OLIVEIRA
Externa à Instituição - MARGARITA MATIAS ROBAINA
Presidente - PAULO RENATO ALVES FIRMINO