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

Timestamp information recorded in event logs plays a crucial role in uncovering meaningful insights into business process performance and behaviour via Process Mining techniques Inaccurate or incomplete timestamps may cause activities in a business process to be ordered incorrectly, leading to unrepresentative process models and incorrect process performance analysis results Thus, the quality of timestamps in an event log should be evaluated thoroughly before the log is used as input for any Process Mining activity To the best of our knowledge, research on the (automated) quality assessment of event logs remains scarce Our work presents an automated approach for detecting and quantifying timestamp-related issues (timestamp imperfections) in an event log We define 15 metrics related to timestamp quality across two axes: four levels of abstraction (event, activity, trace, log) and four quality dimensions (accuracy, completeness, consistency, uniqueness) We adopted the design science research paradigm and drew from knowledge related to data quality as well as event log quality The approach has been implemented as a prototype within the open-source Process Mining framework ProM and evaluated using three real-life event logs and involving experts from practice This approach paves the way for a systematic and interactive enhancement of timestamp imperfections during the data pre-processing phase of Process Mining projects

/pdf/enhancing-event-log-quality-detecting-and-quantifying-53hhltvqq9.pdf

Enhancing event log quality:Detecting and quantifying timestamp imperfections

This paper reports the experiences of faculty in teaching about SAP R/3, teaching IS concepts using SAP R/3 and teaching post-graduate students how to research about implementation issues with ERP.

/pdf/lessons-from-the-field-a-reflection-on-teaching-sap-r-3-and-of560exgz1.pdf

Lessons from the field: A reflection on teaching SAP R/3 and ERP implementation issues

Process mining is, by now, a well-established discipline focussing on process-oriented data analysis. As with other forms of data analysis, the quality and reliability of insights derived through analysis is directly related to the quality of the input (garbage in - garbage out). In the case of process mining, the input is an event log comprised of event data captured (in information systems) during the execution of the process. It is crucial then that the event log be treated as a first-class citizen. While data quality is an easily understood concept little effort has been directed towards systematically detecting data quality issues in event logs. Analysts still spend a large proportion of any project in ‘data cleaning’, often involving manual and ad hoc tasks, and requiring more than one tool. While there are existing tools and languages that query event logs, the problem of different approaches for different log imperfections remains. In this paper we take the first steps to developing QUELI (Querying Event Log for Imperfections) a log query language that provides direct support for detecting log imperfections. We develop an approach that identifies capabilities required of QUELI and illustrate the approach by applying it to 5 of the 11 event log imperfection patterns described in [29]. We view this as a first step towards operationalising systematic, automated support for log cleaning.

Towards event log querying for data quality: Let's start with detecting log imperfections

Process-oriented data mining (process mining) uses algorithms and data (in the form of event logs) to construct models that aim to provide insights into organisational processes. The quality of the data (both form and content) presented to the modeling algorithms is critical to the success of the process mining exercise. Cleaning event logs to address quality issues prior to conducting a process mining analysis is a necessary, but generally tedious and ad hoc task. In this paper we describe a set of data quality issues, distilled from our experiences in conducting process mining analyses, commonly found in process mining event logs or encountered while preparing event logs from raw data sources. We show that patterns are used in a variety of domains as a means for describing commonly encountered problems and solutions. The main contributions of this article are in showing that a patterns-based approach is applicable to documenting commonly encountered event log quality issues, the formulation of a set of components for describing event log quality issues as patterns, and the description of a collection of 11 event log imperfection patterns distilled from our experiences in preparing event logs. We postulate that a systematic approach to using such a pattern repository to identify and repair event log quality issues benefits both the process of preparing an event log and the quality of the resulting event log. The relevance of the pattern-based approach is illustrated via application of the patterns in a case study and through an evaluation by researchers and practitioners in the field.

Event log imperfection patterns for process mining: Towards a systematic approach to cleaning event logs

(a)

Situation faced: Improving Emergency Department (ED) patient flows in terms of processing time, resource use, costs, and patient outcomes is a priority for health service professionals and is vital to the delivery of safe, timely, and effective patient care. Poor patient flows manifest as overcrowding in the ED, prolonged length of stay (LoS), patients “boarding” in EDs and “access block” for admission to inpatient wards. Consequences include poor patient outcomes, reduced access for new patients who present at the ED, and negative effects on staff, including dissatisfaction and stress. Further motivation for improving patient flows in EDs arises because Commonwealth- and state-sponsored financial incentives for hospitals are tied to achieving targets for improved patient access to emergency services. One measure of such improved access is meeting nationally agreed targets for the percentage of patients who are physically discharged from the ED within 4 h of arrival.




(b)

Action taken: A key challenge in deriving evidence-based improvements for patient flows is that of gaining insight into the process factors and context factors that affect patient flows. The case study reported here adopted the BPM Lifecycle reference framework to improve patient flows. In particular we focused on the process identification, discovery, and analysis phases of the BPM Lifecycle. Process-oriented data-mining techniques were applied to real practices to discover models of current patient flows in the ED of St. Andrew’s War Memorial Hospital (SAWMH) in Queensland, Australia. The discovered models were used to evaluate the effect on patient flows of certain context factors of interest to stakeholders. Case histories of 1473 chest pain presentations at SAWMH between September 2011 and March 2013 were analyzed to determine process differences between ED patients with short stays ( 4 h).




(c)

Results achieved: Process models were discovered for the hospital’s ED patient flow. From a control-flow perspective, only minor differences were observed between short- and long-stay patients at SAWMH, although there were timing differences in reaching specific milestone events. Waiting time in the ED following a request for hospital admission added significantly to overall ED LoS.




(d)

Lessons learned: This project demonstrated that process mining is applicable to complex, semi-structured processes like those found in the healthcare domain. Comparative process performance analysis yielded some insights into ED patient flows, including recognition of recurring data-quality issues in datasets extracted from hospital information systems. The templated recognition and resolution of such issues offers a research opportunity to develop a (semi-)automated data-cleaning approach that would alleviate the tedious manual effort required to produce high-quality logs. The project highlighted the importance of hospital information systems collecting both start and end times of activities for proper performance analysis (duration, wait time, bottlenecks). Additions to our process-mining toolset include novel comparative process-performance visualization techniques that highlight the similarities and differences among process cohorts.

Robert Andrews

Papers

Enhancing event log quality:Detecting and quantifying timestamp imperfections

Lessons from the field: A reflection on teaching SAP R/3 and ERP implementation issues

Towards event log querying for data quality: Let's start with detecting log imperfections

Event log imperfection patterns for process mining: Towards a systematic approach to cleaning event logs

Improving patient flows at St. Andrew's War Memorial Hospital's emergency department through process mining