ILSHR Rumor Spreading Model by Combining SIHR and ILSR Models in Complex Networks

Rumor is an important form of social interaction. However, spreading harmful rumors can have a significant negative impact on social welfare. Therefore, it is important to examine rumor models. Rumors are often defined as unconfirmed details or descriptions of public things, events, or issues that are made and promoted through various tools. In this paper, the Ignorant-Lurker-Spreader-Hibernator-Removal (ILSHR) rumor spreading model has been developed by combining the ILSR and SIHR epidemic models. In addition to the characteristics of the lurker group of ILSR, this model also considers the characteristics of the hibernator group of the SIHR model. Due to the complexity of the complex network structure, the state transition function for each node is defined based on their degree to make the proposed model more efficient. Numerical simulations have been performed to compare the ILSHR rumor spreading model with other similar models on the Sina Weibo dataset. The results show more effective ILSHR performance with 95.83% accuracy than CSRT and SIR-IM models.

On the effectiveness of gated echo state networks for data exhibiting long-term dependencies

In the context of recurrent neural networks, gated architectures such as the GRU have contributed to the development of highly accurate machine learning models that can tackle long-term dependencies in the data. However, the training of such networks is performed by the expensive algorithm of gradient descent with backpropagation through time. On the other hand, reservoir computing approaches such as Echo State Networks (ESNs) can produce models that can be trained efficiently thanks to the use of fixed random parameters, but are not ideal for dealing with data presenting long-term dependencies. We explore the problem of employing gated architectures in ESNs from both theoretical and empirical perspectives. We do so by deriving and evaluating a necessary condition for the non-contractivity of the state transition function, which is important to overcome the fading-memory characterization of conventional ESNs. We find that using pure reservoir computing methodologies is not sufficient for effective gating mechanisms, while instead training even only the gates is highly effective in terms of predictive accuracy.

Discrete Optimization on Train Rescheduling on Single-Track Railway: Clustering Hierarchy and Heuristic Search

This paper focuses on discrete dynamic optimization on train rescheduling on single-track railway with the consideration of train punctuality and station satisfaction degree. A discrete dynamic system is firstly described to mimic train rescheduling, and a state transition function is specially designed according to the train departure event. The purpose of this function is to improve simulation efficiency by directly confirming the next discrete time. After the construction and analysis of optimization models to discrete dynamic system, a two-stage heuristic search strategy is developed, by using clustering hierarchy theory and stochastic search strategy, to obtain train departure time and arrival time before each state transition of the system. Finally, a numerical experiment is conducted to verify the proposed system, models, and the heuristic search strategy. The result shows that the discrete dynamic system, together with the state transition function and heuristic search strategy, shows better performance in simulation efficiency and solution quality.

Estimation of Urban Link Travel Time Distribution Using Markov Chains and Bayesian Approaches

Despite the wide application of Floating Car Data (FCD) in urban link travel time and congestion estimation, the sparsity of observations from a low penetration rate of GPS-equipped floating cars make it difficult to estimate travel time distribution (TTD), especially when the travel times may have multimodal distributions that are associated with the underlying traffic states. In this case, the study develops a Bayesian approach based on particle filter framework for link TTD estimation using real-time and historical travel time observations from FCD. First, link travel times are classified by different traffic states according to the levels of vehicle delays. Then, a state-transition function is represented as a Transition Probability Matrix of the Markov chain between upstream and current links with historical observations. Using the state-transition function, an importance distribution is constructed as the summation of historical link TTDs conditional on states weighted by the current link state probabilities. Further, a sampling strategy is developed to address the sparsity problem of observations by selecting the particles with larger weights in terms of the importance distribution and a Gaussian likelihood function. Finally, the current link TTD can be reconstructed by a generic Markov Chain Monte Carlo algorithm incorporating high weighted particles. The proposed approach is evaluated using real-world FCD. The results indicate that the proposed approach provides good accurate estimations, which are very close to the empirical distributions. In addition, the approach with different percentage of floating cars is tested. The results are encouraging, even when multimodal distributions and very few or no observations exist.

Learning and Communicating the Latent States of Human-Machine Collaboration

Artificial agents (both embodied robots and software agents) that interact with humans are increasing at an exceptional rate. Yet, achieving seamless collaboration between artificial agents and humans in the real world remains an active problem. A key challenge is that the agents need to make decisions without complete information about their shared environment and collaborators. For instance, a human-robot team performing a rescue operation after a disaster may not have an accurate map of their surroundings. Even in structured domains, such as manufacturing, a robot might not know the goals or preferences of its human collaborators. Algorithmically, this challenge manifests itself as a problem of decision-making under uncertainty in which the agent has to reason about the latent states of its environment and human collaborator. However, in practice, quantifying this uncertainty (i.e., the state transition function) and even specifying the features (i.e., the relevant states) of human-machine collaboration is difficult. Thus, the objective of this thesis research is to develop novel algorithms that enable artificial agents to learn and reason about the latent states of human-machine collaboration and achieve fluent interaction.

TRANSFER LEARNING FOR CONTINUOUS STATE AND ACTION SPACES

Transfer learning focuses on developing methods to reuse information gathered from a source task in order to improve the learning performance in a related task. In this work, we present a novel approach to transfer knowledge between tasks in a reinforcement learning (RL) framework with continuous states and actions, where the transition and policy functions are approximated by Gaussian processes. The novelty in the proposed approach lies in the idea of transferring information about the hyper-parameters of the state transition function from the source task, which represents qualitative knowledge about the type of transition function that the target task might have, constraining the search space and accelerating the learning process. We performed experiments on relevant tasks for RL, which show a clear improvement in the overall performance when compared to state-of-the-art reinforcement learning and transfer learning algorithms for continuous state and action spaces.

DESIGNING COMPLEX DYNAMICS IN CELLULAR AUTOMATA WITH MEMORY

Since their inception at Macy conferences in later 1940s, complex systems have remained the most controversial topic of interdisciplinary sciences. The term "complex system" is the most vague and liberally used scientific term. Using elementary cellular automata (ECA), and exploiting the CA classification, we demonstrate elusiveness of "complexity" by shifting space-time dynamics of the automata from simple to complex by enriching cells with memory. This way, we can transform any ECA class to another ECA class — without changing skeleton of cell-state transition function — and vice versa by just selecting a right kind of memory. A systematic analysis displays that memory helps "discover" hidden information and behavior on trivial — uniform, periodic, and nontrivial — chaotic, complex — dynamical systems.

Learning-Based Planning

Automated Planning (AP) studies the generation of action sequences for problem solving. A problem in AP is defined by a state-transition function describing the dynamics of the world, the initial state of the world and the goals to be achieved. According to this definition, AP problems seem to be easily tackled by searching for a path in a graph, which is a well-studied problem. However, the graphs resulting from AP problems are so large that explicitly specifying them is not feasible. Thus, different approaches have been tried to address AP problems. Since the mid 90’s, new planning algorithms have enabled the solution of practical-size AP problems. Nevertheless, domain-independent planners still fail in solving complex AP problems, as solving planning tasks is a PSPACE-Complete problem (Bylander, 94). How do humans cope with this planning-inherent complexity? One answer is that our experience allows us to solve problems more quickly; we are endowed with learning skills that help us plan when problems are selected from a stable population. Inspire by this idea, the field of learning-based planning studies the development of AP systems able to modify their performance according to previous experiences. Since the first days, Artificial Intelligence (AI) has been concerned with the problem of Machine Learning (ML). As early as 1959, Arthur L. Samuel developed a prominent program that learned to improve its play in the game of checkers (Samuel, 1959). It is hardly surprising that ML has often been used to make changes in systems that perform tasks associated with AI, such as perception, robot control or AP. This article analyses the diverse ways ML can be used to improve AP processes. First, we review the major AP concepts and summarize the main research done in learning-based planning. Second, we describe current trends in applying ML to AP. Finally, we comment on the next avenues for combining AP and ML and conclude.

On the Equivalence between Small-Step and Big-Step Abstract Machines: A Simple Application of Lightweight Fusion

We show how Ohori and Sasano's recent lightweight fusion by fixed-point promotion provides a simple way to prove the equivalence of the two standard styles of specification of abstract machines: (1) in small-step form, as a state-transition function together with a `driver loop,' i.e., a function implementing the iteration of this transition function; and (2) in big-step form, as a tail-recursive function that directly maps a given configuration to a final state, if any. The equivalence hinges on our observation that for abstract machines, fusing a small-step specification yields a big-step specification. We illustrate this observation here with a recognizer for Dyck words, the CEK machine, and Krivine's machine with call/cc.<br /> <br />The need for such a simple proof is motivated by our current work on small-step abstract machines as obtained by refocusing a function implementing a reduction semantics (a syntactic correspondence), and big-step abstract machines as obtained by CPS-transforming and then defunctionalizing a function implementing a big-step semantics (a functional correspondence).

Refocusing in Reduction Semantics

The evaluation function of a reduction semantics (i.e., a small-step operational semantics with an explicit representation of the reduction context) is canonically defined as the transitive closure of (1) decomposing a term into a reduction context and a redex, (2) contracting this redex, and (3) plugging the contractum in the context. Directly implementing this evaluation function therefore yields an interpreter with a worst-case overhead, for each step, that is linear in the size of the input term. <br /> <br />We present sufficient conditions over the constituents of a reduction semantics to circumvent this overhead, by replacing the composition of (3) plugging and (1) decomposing by a single ``refocus'' function mapping a contractum and a context into a new context and a new redex, if any. We also show how to construct such a refocus function, we prove the correctness of this construction, and we analyze the complexity of the resulting refocus function. <br /> <br />The refocused evaluation function of a reduction semantics implements the transitive closure of the refocus function, i.e., a ``pre-abstract machine.'' Fusing the refocus function with the trampoline function computing the transitive closure gives a state-transition function, i.e., an abstract machine. This abstract machine clearly separates between the transitions implementing the congruence rules of the reduction semantics and the transitions implementing its reduction rules. The construction of a refocus function therefore shows how to mechanically obtain an abstract machine out of a reduction semantics, which was done previously on a case-by-case basis. <br /> <br />We illustrate refocusing by mechanically constructing Felleisen et al.'s CK machine from a call-by-value reduction semantics of the lambda-calculus, and by constructing a substitution-based version of Krivine's machine from a call-by-name reduction semantics of the lambda-calculus. We also mechanically construct three one-pass CPS transformers from three quadratic context-based CPS transformers for the lambda-calculus.