Traditional AI Approaches Research Articles

Unlocking Decision-Making Potential with Causal AI-Driven What-If Simulations

Abstract: This article explores the potential of causal AI-driven what-if simulations in enhancing decision-making across various industries. Causal AI uses theories and causal reasoning to figure out and model the underlying cause-and-effect relationships that govern a system or domain. It can make strong predictions even when it only has limited data. The article discusses the advantages of causal AI over traditional AI approaches and its ability to handle sparse data, enable counterfactual reasoning, and address bias issues. It delves into the applications of what-if simulations powered by causal AI in manufacturing, oil and gas, supply chain management, and contract management, presenting scenarios and demonstrating how causal AI can offer valuable insights and optimize decision-making. The article also highlights the challenges and opportunities associated with causal AI, including the need for domain expertise, integration with existing systems, interpretability and explainability, competitive advantage, positive social impact, research and development, and ethical considerations. The impact of causal AIdriven what-if simulations on decision-making across industries is substantial, enabling organizations to make informed decisions, mitigate risks, and seize opportunities in an ever-changing business landscape.

Real-time detection of electrical load anomalies through hyperdimensional computing

Load anomalies in power distribution systems are relatively rare, yielding imbalanced classification datasets. Consequently, traditional artificial intelligence approaches for detecting load anomalies tend to be relatively inaccurate or rely heavily on pre-processing optimization. Reasoning that hyperdimensional computing (HDC) lends itself naturally to classification with imbalanced datasets, we present an HDC-based method for detecting anomalies accurately and in real time using raw meter data without prior optimization. An associative memory is used to classify the hypervectors, providing resilience against data imbalance. Moreover, a novel retraining function is designed to further improve classification accuracy. The proposed method is evaluated using real-world datasets from two sources with significantly different characteristics—a US university campus that relies on state-of-the-art power monitoring and uses mostly non-renewable energy sources, and a rural village in Tanzania that relies on wireless monitoring and uses 100% solar energy. In comparison with traditional AI approaches, it achieves consistently higher accuracy and runtime efficiency on both datasets and under a wide variety of evaluation metrics. To the best of our knowledge, this work is the first investigation of HDC for anomaly detection in smart grids. Our results show that HDC-based methods have considerable potential for accurate real-time detection of load anomalies.

An expert-model and machine learning hybrid approach to predicting human-agent negotiation outcomes in varied data

We present the results of a machine-learning approach to the analysis of several human-agent negotiation studies. By combining expert knowledge of negotiating behavior compiled over a series of empirical studies with neural networks, we show that a hybrid approach to parameter selection yields promise for designing more effective and socially intelligent agents. Specifically, we show that a deep feedforward neural network using a theory-driven three-parameter model can be effective in predicting negotiation outcomes. Furthermore, it outperforms other expert-designed models that use more parameters, as well as those using other techniques (such as linear regression models or boosted decision trees). In a follow-up study, we show that the most successful models change as the dataset size increases and the prediction targets change, and show that boosted decision trees may not be suitable for the negotiation domain. We anticipate these results will have impact for those seeking to combine extensive domain knowledge with more automated approaches in human-computer negotiation. Further, we show that this approach can be a stepping stone from purely exploratory research to targeted human-behavioral experimentation. Through our approach, areas of social artificial intelligence that have historically benefited from expert knowledge and traditional AI approaches can be combined with more recent proven-effective machine learning algorithms.

Understanding effects of cognitive rehabilitation under a knowledge discovery approach

Traumatic brain injury (TBI) is the leading cause of death and disability in children and young adults worldwide. Cognitive rehabilitation (CR) plans consist of a sequence of CR tasks targeting main cognitive functions. There is not enough on-field experience yet regarding which specific intervention (tasks or exercise assignment) is more appropriate to help therapists to design plans with significant effectiveness on patient improvement. The selection of specific tasks to be prescribed to the patient and the order in which they might be executed is currently decided by the therapists based on their experience.In this paper a new data mining methodology is proposed, combining several tools from Artificial Intelligence, clustering and post-processing analysis to identify regularities in the sequences of tasks in such a way that treatment profiles (classes) can be discovered. Due to the cumulative effect of rehabilitation tasks, small variations within the sequence of tasks performed by the patient do not significantly change the final outcomes in rehabilitation and makes it difficult to find discriminant rules by using the traditional machine learning inductive methods. However, by relaxing the formalization of the problem to find patterns that might include small variations, and introducing motif discovery techniques in the proposed methodology, the complexity of the neurorehabilitation phenomenon can be better captured and a global structure of successful treatment task sequences can be devised.Following this, the relationship between the discovered patterns and the CR treatment response are analyzed, offering a richer perspective than that provided by the single task focus traditionally used in the CR field.The paper provides a definition of the whole methodological approach proposed from a formal point of view, and its application to a real dataset. Comparisons with traditional AI approaches are also presented and the contribution of the proposed methodology to the AI field discussed.

Pathfinding in the cognitive map: Network models of mechanisms for search and planning

Abstract The hippocampus has long been thought to be critical in learning and representing the cognitive map, and thus support functions such as search, pathfinding and route planning. This work aims to demonstrate the utility of hippocampus-based neural networks in modeling human search task behavior. Human solutions to pathfinding problems are generally fast but approximate, in contrast to traditional AI approaches. In this paper, we report data on a human search task, and then examine a set of models, based upon the structure of the hippocampus, which use a goal scent mechanism similar to the optimal pathfinding algorithms used in artificial intelligence systems. We compare five distinct search models, and conclude that a goal scent model driven by multiple goals spread throughout the search space provides the best and most accurate account of the human data. This research suggests a convergence in traditional AI and biologically-inspired approaches to pathfinding that may be mutually beneficial.

An Active Symbols Theory of Chess Intuition

The well-known game of chess has traditionally been modeled in artificial intelligence studies by search engines with advanced pruning techniques. The models were thus centered on an inference engine manipulating passive symbols in the form of tokens. It is beyond doubt, however, that human players do not carry out such processes. Instead, chess masters instead carry out perceptual processes, carefully categorizing the chunks perceived in a position and gradually building complex dynamic structures to represent the subtle pressures embedded in the positions. In this paper we will consider two hypotheses concerning the underlying subcognitive processes and architecture. In the first hypothesis, a multiple-leveled chess representational structure is presented, which includes distance graphs (with varying levels of quality) between pieces, piece mobilities, and abstract roles. These representational schemes seem to account for numerous characteristics of human player's psychology. The second hypothesis concerns the extension of the architecture proposed in the Copycat project as central for modeling the emergent intuitive perception of a chess position. We provide a synthesis on how the postulated architecture models chess intuition as an emergent mixture of simultaneous distance estimations, chunk perceptions, abstract role awareness, and intention activations. This is an alternative model to the traditional AI approaches, focusing on the philosophy of active symbols.