China
Africa
Explore chapters and articles related to this topic
Uploading Consciousness—The Evolution of Conscious Machines of the Future
Published in Anirban Bandyopadhyay, Nanobrain, 2020
The echo state network (ESN) is a recurrent neural network with a sparsely connected hidden layer (with typically 1% connectivity). The connectivity and weights of hidden neurons are randomly assigned and are fixed. Learning the weights of output neurons enables producing specific temporal patterns. Although its behavior is non-linear, the only variables are the weights of the output layer. The error function is thus quadratic with respect to the parameter vector and can be differentiated easily to a linear system.
Time series forecasting for port throughput using recurrent neural network algorithm
Published in Journal of International Maritime Safety, Environmental Affairs, and Shipping, 2021
Nguyen Duy Tan, Hwang Chan Yu, Le Ngoc Bao Long, Sam-Sang You
The recurrent neural network (RNNs) represent a large and varied class of computational models that are designed by more or less detailed analogy with biological brain (neural computing) modules (Lukoševičius and Jaeger 2009). RNN is a class of artificial neural network that permits continuing information related to past knowledge by utilizing a special kind of looped architecture. In fact, the machine learning schemes are employed in many areas regarding data with sequences, such as predicting next word of a sentence. The echo state network (ESN) is a type of reservoir computing (RC) and it shares the basic structure with liquid state machine (LSM) where the internal layers are fixed with random weights, and only the output layer is updated with the weight. They belong to the RNN family and the supervised machine learning principle. Since they are dynamically characterized, they are typically used for learning dynamical processes, modelling of a biological system, signal forecasting and generators.
Sustainability and robust decision-support strategy for multi-echelon supply chain system against disruptions
Published in International Journal of Logistics Research and Applications, 2023
Le Ngoc Bao Long, Truong Ngoc Cuong, Hwan-Seong Kim, Sam-Sang You
The purpose of this paper is to provide answers to those research questions, which give the study a clear focus. For the first question, different qualitative and quantitative tools for nonlinear dynamical analysis such as time series, phase plane, Lyapunov exponents, bifurcation diagram, entropy analysis, and Poincaré map are used to analyze the multi-echelon supply chain model. From a supply chain management perspective, chaotic behaviours also appear in every entity, such as demand, inventory, production, and delivery, as inevitable dynamical characteristics such that every minor change in a link can cause amplified effects on others later on (the BWE). The Hurst exponent computation (HE) will discover the second question, which describes the time-series dataset's predictability. Finally, to resolve the last one, a recurrent neural network (RNN) framework named echo state network (ESN) will be employed to foresee the states of the SCN and to imitate the complex synchronisation effort that is difficult to be characterised analytically. The main contributions of this work are listed as follows: Nonlinear discrete-time dynamics formulate a multi-echelon supply chain model with time delay. Dynamic interactions of components across the supply chain are represented by complex behaviours and investigated by nonlinear data analytics, including Lyapunov exponents, bifurcation graph, permutation entropy, and Poincaré section analysis.With time delay, the bullwhip effect in the supply chain is analyzed by measuring the variance propagation along the upstream when there is a minor difference in demand forecasting.After verifying the predictability with the Hurst exponent, the ESN algorithm is implemented to provide the designated state of the system in the future and to memorise the strategic decisions that have been successfully applied in the past.A robust active controller realised by feedback linearisation is combined with ESN to mitigate supply chain risks caused by undesired factors and recover business activities. The decision support strategy is implemented by synchronising the actual response with the desired behaviour of a well-known system.