Reliable AI Through SVDD and Rule Extraction

Carlevaro, Alberto; Mongelli, Maurizio

doi:10.1007/978-3-030-84060-0_10

Cited by 3 publications

(3 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The new dataset is then elaborated via LLM. Differently from Carlevaro and Mongelli's work, 5 we need a more refined sampling of SVDD classification to derive the new dataset. The sampling is performed by setting a threshold ", such that the extracted observations are sufficiently close to the boundary of the trained and tested SVDD.…”

Section: Rules Extraction From Svddmentioning

confidence: 99%

“…Finally, we report in the following the plot (see Figure 11) concerning the comparison between rule extraction methods with and without the sampling of the points around the edge of the SVDD region (the old algorithm is the one of Carlevaro and Mongelli's work 5 ). It is clear that the accuracy of the classification has been improved with the new version of the ExplainableSVDD algorithm, thus confirming the observations reported so far.…”

Section: Dns Tunnelingmentioning

confidence: 99%

“…Accuracy classification of different extractions of 10 3 subsets of the DNS tunneling dataset. In the legend, the asterisked algorithms at the top ( Ã ) refer to those reported in this article, with the rule extraction near the SVDD edge, while those asterisked at the bottom ( Ã ) refer to the previous version 5. It is clear that the accuracy of the classification is definitely improved by the new approach.…”

mentioning

confidence: 97%

See 2 more Smart Citations

A New SVDD Approach to Reliable and Explainable AI

Carlevaro

Mongelli

2022

IEEE Intell. Syst.

Self Cite

View full text Add to dashboard Cite

Section: Rules Extraction From Svddmentioning

confidence: 99%

Section: Dns Tunnelingmentioning

confidence: 99%

mentioning

confidence: 97%

See 1 more Smart Citation

A New SVDD Approach to Reliable and Explainable AI

Carlevaro

Mongelli

2022

IEEE Intell. Syst.

Self Cite

View full text Add to dashboard Cite

eXplainable and Reliable Against Adversarial Machine Learning in Data Analytics

et al. 2022

Self Cite

View full text Add to dashboard Cite

Machine learning (ML) algorithms are nowadays widely adopted in different contexts to perform autonomous decisions and predictions. Due to the high volume of data shared in the recent years, ML algorithms are more accurate and reliable since training and testing phases are more precise. An important concept to analyze when defining ML algorithms concerns adversarial machine learning attacks. These attacks aim to create manipulated datasets to mislead ML algorithm decisions. In this work, we propose new approaches able to detect and mitigate malicious adversarial machine learning attacks against a ML system. In particular, we investigate the Carlini-Wagner (CW), the fast gradient sign method (FGSM) and the Jacobian based saliency map (JSMA) attacks. The aim of this work is to exploit detection algorithms as countermeasures to these attacks. Initially, we performed some tests by using canonical ML algorithms with a hyperparameters optimization to improve metrics. Then, we adopt original reliable AI algorithms, either based on eXplainable AI (Logic Learning Machine) or Support Vector Data Description (SVDD). The obtained results show how the classical algorithms may fail to identify an adversarial attack, while the reliable AI methodologies are more prone to correctly detect a possible adversarial machine learning attack. The evaluation of the proposed methodology was carried out in terms of good balance between FPR and FNR on real world application datasets: Domain Name System (DNS) tunneling, Vehicle Platooning and Remaining Useful Life (RUL). In addition, a statistical analysis was performed to improve the robustness of the trained models, including evaluating their performance in terms of runtime and memory consumption. INDEX TERMS machine learning, detection algorithms, adversarial machine learning, reliable I. INTRODUCTION A. BACKGROUND VOLUME 4, 2016

show abstract

A novel method to derive personalized minimum viable recommendations for type 2 diabetes prevention based on counterfactual explanations

et al. 2022

Self Cite

View full text Add to dashboard Cite

Despite the growing availability of artificial intelligence models for predicting type 2 diabetes, there is still a lack of personalized approaches to quantify minimum viable changes in biomarkers that may help reduce the individual risk of developing disease. The aim of this article is to develop a new method, based on counterfactual explanations, to generate personalized recommendations to reduce the one-year risk of type 2 diabetes. Ten routinely collected biomarkers extracted from Electronic Medical Records of 2791 patients at low risk and 2791 patients at high risk of type 2 diabetes were analyzed. Two regions characterizing the two classes of patients were estimated using a Support Vector Data Description classifier. Counterfactual explanations (i.e., minimal changes in input features able to change the risk class) were generated for patients at high risk and evaluated using performance metrics (availability, validity, actionability, similarity, and discriminative power) and a qualitative survey administered to seven expert clinicians. Results showed that, on average, the requested minimum viable changes implied a significant reduction of fasting blood sugar, systolic blood pressure, and triglycerides and a significant increase of high-density lipoprotein in patients at risk of diabetes. A significant reduction in body mass index was also recommended in most of the patients at risk, except in females without hypertension. In general, greater changes were recommended in hypertensive patients compared to non-hypertensive ones. The experts were overall satisfied with the proposed approach although in some cases the proposed recommendations were deemed insufficient to reduce the risk in a clinically meaningful way. Future research will focus on a larger set of biomarkers and different comorbidities, also incorporating clinical guidelines whenever possible. Development of additional mathematical and clinical validation approaches will also be of paramount importance.

show abstract

Reliable AI Through SVDD and Rule Extraction

Cited by 3 publications

References 25 publications

A New SVDD Approach to Reliable and Explainable AI

A New SVDD Approach to Reliable and Explainable AI

eXplainable and Reliable Against Adversarial Machine Learning in Data Analytics

A novel method to derive personalized minimum viable recommendations for type 2 diabetes prevention based on counterfactual explanations

Contact Info

Product

Resources

About