Integration of Neural Networks and knowledge-based systems in medicine

Ultsch, Alfred; Korus, Dieter; Kleine, T. O.

doi:10.1007/3-540-60025-6_170

Cited by 15 publications

(4 citation statements)

References 2 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the sense of the latter, explanation represents a distinct approach to extract information from the learned model [21]. Typical interpretable ML systems consist of combinations of neural networks and rule-based expert systems [22,23], classification based on predictive association rules [24], Bayesian networks with rule mining [25], hybrids of clustering and fuzzy classification [26] or neuro-fuzzy classification [27], non-iterative ANN-based XAIs [28,29], rule lists [30], interpretable decision sets [31] or decision tables [32], decision tree clustering [33,34] or clustering combined with generative models [35]. The two most recent XAI approaches are the unsupervised decision tree clustering eUD3.5 [36] and a hybrid of k-means clustering and a top-down decision tree [37].…”

Section: Related Workmentioning

confidence: 99%

Explainable AI Framework for Multivariate Hydrochemical Time Series

Thrun

Ultsch

2021

MAKE

View full text Add to dashboard Cite

The understanding of water quality and its underlying processes is important for the protection of aquatic environments. With the rare opportunity of access to a domain expert, an explainable AI (XAI) framework is proposed that is applicable to multivariate time series. The XAI provides explanations that are interpretable by domain experts. In three steps, it combines a data-driven choice of a distance measure with supervised decision trees guided by projection-based clustering. The multivariate time series consists of water quality measurements, including nitrate, electrical conductivity, and twelve other environmental parameters. The relationships between water quality and the environmental parameters are investigated by identifying similar days within a cluster and dissimilar days between clusters. The framework, called DDS-XAI, does not depend on prior knowledge about data structure, and its explanations are tendentially contrastive. The relationships in the data can be visualized by a topographic map representing high-dimensional structures. Two state of the art XAIs called eUD3.5 and iterative mistake minimization (IMM) were unable to provide meaningful and relevant explanations from the three multivariate time series data. The DDS-XAI framework can be swiftly applied to new data. Open-source code in R for all steps of the XAI framework is provided and the steps are structured application-oriented.

show abstract

Section: Related Workmentioning

confidence: 99%

Explainable AI Framework for Multivariate Hydrochemical Time Series

Thrun

Ultsch

2021

MAKE

View full text Add to dashboard Cite

show abstract

“…In the sense of the latter, explanation represents a distinct approach to extract information from the learned model [17]. Typical interpretable ML systems or so-called XAIs comprise combinations of neural networks with rule-based expert systems [18,19], Bayesian networks with rule mining [20], hybrids of clustering and fuzzy classification [21] or neuro-fuzzy classification [22], interpretable decision sets [23] or decision tables [24], decision tree clustering [25] or clustering combined with generative models [26]. Two of the most recent XAI approaches are the unsupervised decision tree clustering eUD3.5 [27] and a hybrid of k-means clustering and a top-down decision tree [28].…”

Section: Related Workmentioning

confidence: 99%

Explainable AI Framework for Multivariate Hydrochemical Time Series

Thrun¹,

Ultsch²

2020

Preprint

View full text Add to dashboard Cite

The understanding of water quality and its underlying processes is important for the protection of aquatic environments enabling the rare opportunity of access to a domain expert. Hence, an explainable AI (XAI) framework is proposed that is applicable to multivariate time series resulting in explanations that are interpretable by a domain expert. The XAI combines in three steps a data-driven choice of a distance measure with explainable cluster analysis through supervised decision trees. The multivariate time series consists of water quality measurements, including nitrate, electrical conductivity, and twelve other environmental parameters. The relationships between water quality and the environmental parameters are investigated by identifying similar days within a cluster and dissimilar days between clusters. The XAI does not depend on prior knowledge about data structure, and its explanations are tendentially contrastive. The relationships in the data can be visualized by a topographic map representing high-dimensional structures. Two comparable decision-based XAIs were unable to provide meaningful and relevant explanations from the multivariate time series data. Open-source code in R for the three steps of the XAI framework is provided.

show abstract

“…They are non-linear statistical data modeling techniques in which interconnected elements process the information simultaneously, having the ability to adapt and learn from past examples. Artificial Neural Networks have application in various fields such as: pattern recognition; medical diagnosis (Ultsch et al (1995) used the capacity of neural network to diagnose acidosis diseases; Zhou et al (2001) proposed an Artificial Neural Network ensemble in the process of lung cancer diagnosis; Kiyan and Yildirim (2003) used ANNs for breast cancer diagnosis); credit risk models (Amir (2001) developed a Neural Network bankruptcy prediction model; Baesens et al (2003) used Neural Network rule extraction for credit risk evaluation); market prices changes (Yoon and Swales (1993) analyzed the forecasting power of Neural Networks for stock prices) etc.…”

Section: Artificial Neural Networkmentioning

confidence: 99%

Customer Segmentation in Private Banking Sector Using Machine Learning Techniques

Smeureanu

Ruxanda

Badea

2013

Journal of Business Economics and Management

View full text Add to dashboard Cite

Machine learning techniques have proven good performance in classification matters of all kinds: medical diagnosis, character recognition, credit default and fraud prediction, and also foreign exchange market prognosis. Customer segmentation in private banking sector is an important step for profitable business development, enabling financial institutions to address their products and services to homogeneous classes of customers. This paper approaches two of the most popular machine learning techniques, Neural Networks and Support Vector Machines, and describes how each of these perform in a segmentation process.

show abstract

Integration of Neural Networks and knowledge-based systems in medicine

Cited by 15 publications

References 2 publications

Explainable AI Framework for Multivariate Hydrochemical Time Series

Explainable AI Framework for Multivariate Hydrochemical Time Series

Explainable AI Framework for Multivariate Hydrochemical Time Series

Customer Segmentation in Private Banking Sector Using Machine Learning Techniques

Contact Info

Product

Resources

About