| Abstrakt: |
Russia's unjustified and unprovoked invasion of Ukraine has created shock waves in global energy markets, leading to price economic uncertainty, supply shortages, volatility and security issues. On the other hand, the economic disruption caused by the war has amplified calls for an accelerated energy transition. In this paper, we first proceed to an exploratory data analysis of the global energy market and, in particular, of the Romanian electricity market, with an emphasis on the effects of the war on the market trend and structure. Next, we used available data to build nonlinear autoregressive models with exogenous inputs (NARX), based on deep learning neural network architectures (LSTM: Long Short-Term Memory and CNN: Convolutional Neural Networks), for predicting certain endogenous variables of energy market depending on some chosen exogenous inputs. [ABSTRACT FROM AUTHOR] |
