Towards a Consistent Interpretation of AIOps Models

Artificial Intelligence for IT Operations (AIOps) has been adopted in organizations in various tasks, including interpreting models to identify indicators of service failures. To avoid misleading practitioners, AIOps model interpretations should be consistent (i.e., different AIOps models on the same task agree with one another on feature importance). However, many AIOps studies violate established practices in the machine learning community when deriving interpretations, such as interpreting models with suboptimal performance, though the impact of such violations on the interpretation consistency has not been studied. In this article, we investigate the consistency of AIOps model interpretation along three dimensions: internal consistency, external consistency, and time consistency. We conduct a case study on two AIOps tasks: predicting Google cluster job failures and Backblaze hard drive failures. We find that the randomness from learners, hyperparameter tuning, and data sampling should be controlled to generate consistent interpretations. AIOps models with AUCs greater than 0.75 yield more consistent interpretation compared to low-performing models. Finally, AIOps models that are constructed with the Sliding Window or Full History approaches have the most consistent interpretation with the trends presented in the entire datasets. Our study provides valuable guidelines for practitioners to derive consistent AIOps model interpretation.

Validation of the Romanian Version of the Toronto Empathy Questionnaire (TEQ) among Undergraduate Medical Students

Medical professionals require adequate abilities to identify others’ emotions and express personal emotions. We aimed to determine the validity and reliability of an empathy measuring tool in medical students for this study. We employed Spreng’s Toronto Empathy Questionnaire (TEQ) as a starting point for this validation. The process was performed in several steps, including an English-Romanian-English translation and a focus group meeting to establish each question’s degree of understandability and usability, with minor improvements of wording in each step. We checked internal and external consistency in a pilot group (n = 67). For construct and convergent validity, we used a sample of 649 students. The overall internal and external reliability performed well, with Cronbach’s alpha = 0.727 and respective ICC = 0.776. The principal component analysis resulted in 3 components: prosocial helping behavior, inappropriate sensitivity, dismissive attitude. Component 1 includes positively worded questions, and components 2 and 3 include negatively worded questions. Women had significantly higher scores than men in convergent validity, but we did not highlight any differences for other demographic factors. The Romanian version of the TEQ is a reliable and valid tool to measure empathy among undergraduate medical students that may be further used in subsequent research.

Criminalization of drunk driving: problems of legislative practice

The article highlights the Ukrainian experience of criminalization of drunk driving as a result of amendments to criminal and administrative legislation in 2018–2021. The importance of systematization and validity in making changes to administrative and criminal law is noted. Attention is drawn to the need for careful observance of the rules of legal technique in legislative activity, given that the use of administrative and criminal law is associated with the most significant restrictions on human rights and freedoms. It is noted that there are violations of the rules of legal technique, allowed in the relevant laws in terms of the provisions on criminalization and decriminalization of drunk driving: violations of the homogeneity of legal regulation (Law № 720-IX regulated an issue that was not the subject of its regulation); internal consistency (Law № 720-IX on amendments “in connection with the adoption of Law № 2617-VIII” amended the Law itself № 2617-VIII); external consistency (provisions of Law № 720-IX contradict the provisions of Article 2 of the Code of Administrative Offenses and Article 3 of the Criminal Code of Ukraine); linguistic (in paragraph 117 of the Law № 720-IX there is a morphological error); procedural (violated the requirements of Articles 90, 92 of the Regulations). Attention is drawn to the content of the conclusions and the legal significance of the explanations of the Parliamentary Committee on Law Enforcement, adopted on the criminalization of drunk driving. It is noted that the relevant committee violated the regulatory procedures and provisions of the legislation on parliamentary committees. The consequences of the relevant technical and legal violations (legislative uncertainty) are highlighted and ways to eliminate these problems are suggested.

Associations of Disease-Modifying Therapies With COVID-19 Severity in Multiple Sclerosis

Background:People with multiple sclerosis (MS) are a vulnerable group for severe COVID- 19, particularly those taking immunosuppressive disease-modifying therapies (DMTs). We examined the characteristics of COVID-19 severity in an international sample of people with MS.Methods:Data from 12 data-sources in 28 countries were aggregated (sources could include patients from 1-12 countries). Demographic (age, sex), clinical (MS-phenotype, disability), and DMT (untreated, alemtuzumab, cladribine, dimethyl-fumarate, glatiramer acetate, interferon, natalizumab, ocrelizumab, rituximab, siponimod, other DMTs) covariates were queried, alongside COVID-19 severity outcomes, hospitalisation, ICU admission, requiring artificial ventilation, and death. Characteristics of outcomes were assessed in patients with suspected/confirmed COVID-19 using multilevel mixed-effects logistic regression, adjusted for age, sex, MS-phenotype, and EDSS.Results:657(28.1%) with suspected and 1,683(61.9%) with confirmed COVID-19 were analysed. Among suspected+confirmed and confirmed-only COVID-19, 20.9% and 26.9% were hospitalised, 5.4% and 7.2% were admitted to ICU, 4.1% and 5.4% required artificial ventilation, and 3.2% and 3.9% died. Older age, progressive MS-phenotype, and higher disability were associated with worse COVID-19 outcomes. Compared to dimethyl-fumarate, ocrelizumab and rituximab were associated with hospitalisation (aOR=1.56,95%CI=1.01- 2.41; aOR=2.43,95%CI=1.48-4.02) and ICU admission (aOR=2.30,95%CI=0.98-5.39;aOR=3.93,95%CI=1.56-9.89), though only rituximab was associated with higher risk of artificial ventilation (aOR=4.00,95%CI=1.54-10.39). Compared to pooled other DMTs, ocrelizumab and rituximab were associated with hospitalisation (aOR=1.75,95%CI=1.29- 2.38; aOR=2.76,95%CI=1.87-4.07) and ICU admission (aOR=2.55,95%CI=1.49-4.36;aOR=4.32,95%CI=2.27-8.23) but only rituximab with artificial ventilation (aOR=6.15,95%CI=3.09-12.27). Compared to natalizumab, ocrelizumab and rituximab wereassociated with hospitalisation (aOR=1.86,95%CI=1.13-3.07; aOR=2.88,95%CI=1.68-4.92) and ICU admission (aOR=2.13,95%CI=0.85-5.35; aOR=3.23,95%CI=1.17-8.91), but only rituximab with ventilation (aOR=5.52,95%CI=1.71-17.84). Importantly, associations persisted on restriction to confirmed COVID-19 cases. No associations were observed between DMTs and death. Stratification by age, MS-phenotype, and EDSS found no indications that DMT associations with COVID-19 severity reflected differential DMT allocation by underlying COVID-19 severity.Conclusions:Using the largest cohort of people with MS and COVID-19 available, we demonstrated consistent associations of rituximab with increased risk of hospitalisation, ICU admission, and requiring artificial ventilation, and ocrelizumab with hospitalisation and ICU admission. Despite the study’s cross-sectional design, the internal and external consistency of these results with prior studies suggests rituximab/ocrelizumab use may be a risk factor for more severe COVID-19.

1298Associations of DMT therapies with COVID-19 severity in multiple sclerosis

Background People with multiple sclerosis (MS) are a vulnerable group for severe COVID-19, particularly those taking immunosuppressive disease-modifying therapies (DMTs). We examined the characteristics of COVID-19 severity in an international sample of people with MS. Methods Data from 12 data-sources in 28 countries were aggregated (sources could include patients from 1-12 countries). Demographic (age, sex), clinical (MS phenotype, disability), and DMT (untreated, alemtuzumab, cladribine, dimethyl-fumarate, glatiramer-acetate, interferon, natalizumab, ocrelizumab, rituximab, siponimod, other) covariates were queried, alongside COVID-19 hospitalisation, admission to ICU, requiring artificial ventilation, and death. Characteristics of outcomes were assessed in patients with suspected/confirmed COVID-19 using multilevel mixed-effects logistic regression, adjusted for age, sex, MS phenotype, and EDSS. Results 657 (28.1%) with suspected and 1,683 (61.9%) with confirmed COVID-19 were analysed. Among suspected+confirmed/confirmed-only COVID-19, 20.9%/26.9% were hospitalised, 5.4%/7.2% were admitted to ICU, 4.1%/5.4% required artificial ventilation, and 3.2%/3.9% died. Older age, progressive MS-phenotype, and higher disability were associated with worse COVID-19 outcomes. Compared to dimethyl-fumarate, ocrelizumab and rituximab were associated with hospitalisation (aOR=1.56,95%CI=1.01-2.41; aOR=2.43,95%CI=1.48-4.02) and ICU admission (aOR=2.30,95%CI=0.98-5.39; aOR=3.93,95%CI=1.56-9.89), though only rituximab was associated with higher risk of artificial ventilation (aOR=4.00,95%CI=1.54-10.39). Importantly, associations persisted on restriction to confirmed COVID-19 cases. No associations were observed between DMTs and death. Conclusions Despite the cross-sectional design of this study, the internal and external consistency of these results with prior studies suggests their use may be a risk factor for more severe COVID-19. Key messages Anti-CD20 DMTs may be associated with worse COVID-19 severity amongst people with multiple sclerosis.

Longino's Concept of Values in Science

While classical neo-positivists reject any role for traditionally understood values in science, Kuhn identifies five specific values as criteria for assessing a scientific theory; this approach has been further developed by several other authors. This paper focuses on Helen Longino, who presents a significant contemporary critique of Kuhn’s concept. The most controversial aspect of Longino’s position is arguably her claim that the criterion of empirical adequacy is the least defensible basis for assessing theories. The de-emphasizing of the importance of external consistency as a value and the introduction of socio-political considerations into the processes of an assessment of scientific theories are also considered problematic issues. I provide arguments against Longino’s conception, identify some of its problems, and argue for refusal of her approach.

Large-Scale Assessment of Health-Related Physical Fitness in French Older Adults: Feasibility and Validity

Objectives: This study aims to assess the validity, internal consistency, implementation, and feasibility of a sequence of tests, the Vitality Test Battery, designed to measure physical fitness, at a large scale in French older adults.Methods: A total of 528 volunteers (age ≥60 years) took the battery of 10 tests: 6-min walk, trunk strength, hand grip strength, medicine ball throwing, 30-s chair stand, flexibility, balance, plate tapping, ruler drop, and dual task.Results: Internal consistency was high, with the Cronbach alpha coefficients at around 0.77, explaining 64% of the variance. The test–retest correlations (0.3–0.6) between the items were acceptable and displayed an internal consistency property. Although five components explained 65% of the variance, all the items were kept because their eigenvalues were near to 0.9. External consistency was validated by a significant decrease in fitness scores (p < 0.001) with age and body mass index.Discussion: The Vitality Test Battery is a safe, valid tool for assessing physical fitness in persons aged over 60 years.

Mario Bunge’s systemic thesis of truth: implications for research practice and the “reproducibility crisis”

There are currently serious concerns that published scientific findings often fail to be reproducible, and that some solutions may be gleaned by attending the several methodological and sociological recommendations that could be found in the literature. However, researchers would also arrive at some answers by considering the advice of the philosophy of science, particularly semantics, about theses on truth related to scientific realism. Sometimes scientists understand the correspondence thesis of truth (CTT) as asserting that the next unique empirical confirmation of a hypothesis suffices to attribute truth to it provisionally. Such empiricist bias is not necessarily at the core of CTT, but Mario Bunge proposed the synthetic thesis of truth (STT), based on CTT, to explicitly avoid the bias. STT requires considering a hypothesis corroborated, both by purely empirical confirmation and external consistency or compatibility with the bulk of existing background knowledge (systemicity). While a capricious understanding of CTT could be rigged to recommend the “one shot game” in hypothesis testing, STT clearly demands the use of multiple approaches, empirical as well as theoretical, and it asserts that a scientific test is effective to the extent to which it is neither purely empirical, nor viewed in isolation. Pattern consistency (empirical control) together with an understanding of causal relations (rational together with empirical control) make confirmed hypotheses robust and more reliable. The militancy of the double mechanism of hypothesis control of STT can help mitigate the reproducibility crisis. Earl Werner’s research program in community ecology is an example of the (implicit) use of STT criteria, which leads to the development of reliable, cross-checked, ecological results, with high predictive capacity.