https://www2.econ.iastate.edu/tesfatsi/DeepLearningInNeuralNetworksOverview.JSchmidhuber2015.pdf

Deep learning in neural networks

Explainable Artificial Intelligence (XAI): Concepts, Taxonomies, Opportunities and Challenges toward Responsible AI

In recent years, many accurate decision support systems have been constructed as black boxes, that is as systems that hide their internal logic to the user. This lack of explanation constitutes both a practical and an ethical issue. The literature reports many approaches aimed at overcoming this crucial weakness, sometimes at the cost of sacrificing accuracy for interpretability. The applications in which black box decision systems can be used are various, and each approach is typically developed to provide a solution for a specific problem and, as a consequence, it explicitly or implicitly delineates its own definition of interpretability and explanation. The aim of this article is to provide a classification of the main problems addressed in the literature with respect to the notion of explanation and the type of black box system. Given a problem definition, a black box type, and a desired explanation, this survey should help the researcher to find the proposals more useful for his own work. The proposed classification of approaches to open black box models should also be useful for putting the many research open questions in perspective.

https://dl.acm.org/doi/pdf/10.1145/3236009

A Survey of Methods for Explaining Black Box Models

At the dawn of the fourth industrial revolution, we are witnessing a fast and widespread adoption of artificial intelligence (AI) in our daily life, which contributes to accelerating the shift towards a more algorithmic society. However, even with such unprecedented advancements, a key impediment to the use of AI-based systems is that they often lack transparency. Indeed, the black-box nature of these systems allows powerful predictions, but it cannot be directly explained. This issue has triggered a new debate on explainable AI (XAI). A research field holds substantial promise for improving trust and transparency of AI-based systems. It is recognized as the sine qua non for AI to continue making steady progress without disruption. This survey provides an entry point for interested researchers and practitioners to learn key aspects of the young and rapidly growing body of research related to XAI. Through the lens of the literature, we review the existing approaches regarding the topic, discuss trends surrounding its sphere, and present major research trajectories.

Peeking Inside the Black-Box: A Survey on Explainable Artificial Intelligence (XAI)

Artificial Neural Networks (ANNs) are being used increasingly to predict and forecast water resources variables. In this paper, the steps that should be followed in the development of such models are outlined. These include the choice of performance criteria, the division and pre-processing of the available data, the determination of appropriate model inputs and network architecture, optimisation of the connection weights (training) and model validation. The options available to modellers at each of these steps are discussed and the issues that should be considered are highlighted. A review of 43 papers dealing with the use of neural network models for the prediction and forecasting of water resources variables is undertaken in terms of the modelling process adopted. In all but two of the papers reviewed, feedforward networks are used. The vast majority of these networks are trained using the backpropagation algorithm. Issues in relation to the optimal division of the available data, data pre-processing and the choice of appropriate model inputs are seldom considered. In addition, the process of choosing appropriate stopping criteria and optimising network geometry and internal network parameters is generally described poorly or carried out inadequately. All of the above factors can result in non-optimal model performance and an inability to draw meaningful comparisons between different models. Future research efforts should be directed towards the development of guidelines which assist with the development of ANN models and the choice of when ANNs should be used in preference to alternative approaches, the assessment of methods for extracting the knowledge that is contained in the connection weights of trained ANNs and the incorporation of uncertainty into ANN models.

Neural networks for the prediction and forecasting of water resources variables: a review of modelling issues and applications

In this paper we report on key findings and lessons from a process mining case study conducted to analyse transport pathways discovered across the time-critical phase of pre-hospital care for persons involved in road traffic crashes in Queensland (Australia). In this study, a case is defined as being an individual patient’s journey from roadside to definitive care. We describe challenges in constructing an event log from source data provided by emergency services and hospitals, including record linkage (no standard patient identifier), and constructing a unified view of response, retrieval, transport and pre-hospital care from interleaving processes of the individual service providers. We analyse three separate cohorts of patients according to their degree of interaction with Queensland Health’s hospital system (C1:no transport required, C2:transported but no Queensland Health hospital, C3:transported and hospitalisation). Variant analysis and subsequent process modelling show high levels of variance in each cohort resulting from a combination of data collection, data linkage and actual differences in process execution. For Cohort 3, automated process modelling generated ‘spaghetti’ models. Expert-guided editing resulted in readable models with acceptable fitness, which were used for process analysis. We also conduct a comparative performance analysis of transport segment based on hospital ‘remoteness’. With regard to the field of process mining, we reach various conclusions including (i) in a complex domain, the current crop of automated process algorithms do not generate readable models, however, (ii) such models provide a starting point for expert-guided editing of models (where the tool allows) which can yield models that have acceptable quality and are readable by domain experts, (iii) process improvement opportunities were largely suggested by domain experts (after reviewing analysis results) rather than being directly derived by process mining tools, meaning that the field needs to become more prescriptive (automated derivation of improvement opportunities).

https://www.mdpi.com/1660-4601/17/10/3426/pdf

A Comparative Process Mining Analysis of Road Trauma Patient Pathways.

Context - Healthcare processes such as patient flows in the emergency department, are notoriously complex because of their high degree of variability. Yet, hospital stakeholders need to understand differences between variants of these processes, e.g. between different patient cohorts or different times of the day, to offer a consistent and reliable service to their patients.

Objective

 - Analysis of these processes through classic process mining techniques is hampered by the inherently high variability of these processes.

Method

 - This paper contributes two novel visualisation and animation techniques, specifically catered for highly-varied business processes, which can distill both static and dynamic differences between variants of the same business process, using their execution logs.

Results

 - We validate these techniques via two case studies with two Australian hospitals and obtained insights on the differences that underpin patient flows for various patient cohorts through the Emergency Department.

Conclusion

 - The Configurable Model Visualisation technique provides a birds-eye view of similarities and differences be- tween process variants while the Log Animation tool addresses behavioural comparison and is unique in its ability to display concurrently the processing of multiple process variants through a single BMPN model.

Comparing static and dynamic aspects of patient flows via process model visualisations

This paper describes RULEIN/RULEX; an automated technique for the refinement of a knowledge base. RULEIN constructs a Rapid Backprop (RBP) network from an initial, partially complete/accurate knowledge base which is formulated as a set of propositional rules. The RBP network is then trained on a set of examples drawn from the problem domain. RULEX is then applied to the weights of the trained network to extract a set of !refined propositional rules. The refined rule set represents the original knowledge base modified in the light of network training. Network training has the potential to remove inaccuracies in the original rule base, supplement partially correct initial rules, and add new rules. Rule initialisation can also speed up network learning by obviating the necessity of starting training from a tabula rasa configuration. RULEIN/RULEX is evaluated using rule !quality criteria and results are presented for some benchmark problems. The method has application in many areas but is particularly suited to overcoming the so called !knowledge acquisition bottleneck in the knowledge engineering phase of rule based expert system construction.

Refining Expert Knowledge with an Artificial Neural Network.

This paper quantitatively examines the effects of initialising a Rapid Backprop Network (REP) with prior domain knowledge expressed in the form of propositional rules. The paper first describes the RBP network and then introduces the RULEIN algorithm which encodes propositional rules as the weights of the nodes of the REP network. A selection of datasets is used to compare networks that began learning from tabula rasa with those that were initialised with varying amounts of domain knowledge prior to the commencement of the learning phase. Network performance is compared in terms of time to converge, accuracy at convergence, and network size at convergence.

On the effects of initialising a neural network with prior knowledge

Process mining provides analytical tools and methods which can distil insights about process behaviour from big process-related data. Yet challenges relating to the impact of poor quality data on event logs, the input to process mining analyses, remain. Despite researchers raising concerns about event log data quality, event log preparation is, in practice, generally handled mechanistically, focusing on fixing symptoms rather than on uncovering the root causes of event log data quality issues. To address this, we introduce the Odigos (Greek for “guide”) framework. Based on semiotics and Peircean abductive reasoning, the Odigos framework facilitates an informed way of dealing with data quality issues in event logs. Odigos supports both prognostic (foreshadowing potential quality issues) and diagnostic (identifying root causes of discovered quality issues) approaches. We examine in depth how the framework supports a detailed root-cause analysis of a well-known collection of event log imperfection patterns.

Robert Andrews

Papers

A Comparative Process Mining Analysis of Road Trauma Patient Pathways.

Comparing static and dynamic aspects of patient flows via process model visualisations

Refining Expert Knowledge with an Artificial Neural Network.

On the effects of initialising a neural network with prior knowledge

Root-cause analysis of process-data quality problems