Search

The fairness of machine learning-based decision support systems has become a critical issue, also in the field of predicting the success of venture capital investment startups. Inappropriate allocation of venture capital, fueled by discriminatory biases, can lead to missed investment opportunities and poor investment decisions. Despite numerous studies that have addressed the prevalence of biases in venture capital allocation and decision support models, few have addressed the importance of incorporating fairness into the modeling process. In this study, we leverage invariant feature representation learning to develop a startup success prediction model using Crunchbase data, while satisfying group fairness. Our results show that discriminatory bias can be significantly reduced with minimal impact on model performance. Additionally, we demonstrate the versatility of our approach by mitigating multiple biases simultaneously. This work highlights the significance of addressing fairness in decisionsupport models to ensure equitable outcomes in venture capital investments.

(1) Background: Mortality is a major outcome in research on chronic obstructive pulmonary disease (COPD) with various predictors described. However, the dynamic courses of important predictors over time are disregarded. This study evaluates if longitudinal assessment of predictors provides additional information on the mortality risk in COPD when compared with a cross-sectional analysis.; (2) In a longitudinal, prospective, non-interventional cohort study including mild to very severe COPD patients, mortality and its various possible predictors were annually assessed up to seven years.; (3) Results: 297 patients were analysed. Mean (SD) age was 62.5 (7.6) years and 66% males. Mean (SD) FEV1 was 48.8 (21.4)%. A total of 105 events (35.4%) happened with a median (95% CI) survival time of 8.2 (7.2/NA) years. No evidence for a difference between the raw variable and the variable history on the predictive value for all tested variables over each visit was found. There was no evidence for changing effect estimates (coefficients) across the study visits due to the longitudinal assessment; (4) Conclusions: We found no evidence that predictors of mortality in COPD are time dependent. This implies that cross-sectional measured predictors show robust effect estimates over time and multiple assessments seem not to change the predictive value of the measure.

Richard Kraemer,1,2,* Fabian Gardin,3,* Hans-Jürgen Smith,4 Florent Baty,5 Jürg Barandun,3 Andreas Piecyk,3 Stefan Minder,1 Jörg Salomon,1 Martin Frey,6 Martin H Brutsche,5 Heinrich Matthys7 1Centre of Pulmonary Medicine, Hirslanden Private Hospital Group, Salem-Hospital, Bern, Switzerland; 2Center for Translational Medicine and Biomedical Entrepreneurship, University of Bern, Bern, Switzerland; 3Centre of Pulmonary Medicine, Hirslanden Private Hospital Group, Clinic Hirslanden, Zürich, Switzerland; 4Medical Development, Research in Respiratory Diagnostics, Berlin, Germany; 5Department of Pneumology, Cantonal Hospital St, Gallen, Switzerland; 6Department of Pneumology, Barmelweid Hospital, Barmelweid, Switzerland; 7Department of Pneumology, University Hospital of Freiburg, Freiburg, Germany*These authors contributed equally to this workCorrespondence: Richard Kraemer, Center of Pulmonary Medicine, Hirslanden Private Hospital Group, Schänzlistrasse 39, Berne, CH-3013, Switzerland, Tel +41 79 300 26 53, Email richard.kraemer@hirslanden.chBackground: A significant proportion of patients with obstructive lung disease have clinical and functional features of both asthma and chronic obstructive pulmonary disease (COPD), referred to as the asthma–COPD overlap (ACO). The distinction of these phenotypes, however, is not yet well-established due to the lack of defining clinical and/or functional criteria. The aim of our investigations was to assess the discriminating power of various lung function parameters on the assessment of ACO.Methods: From databases of 4 pulmonary centers, a total of 540 patients (231 males, 309 females), including 372 patients with asthma, 77 patients with ACO and 91 patients with COPD, were retrospectively collected, and gradients among combinations of explanatory variables of spirometric (FEV1, FEV1/FVC, FEF25-75), plethysmographic (sReff, sGeff, the aerodynamic work of breathing at rest; sWOB), static lung volumes, in

Background and objective: Whether immunological biomarkers combined with clinical characteristics measured during an exacerbation-free period are predictive of acute exacerbation of chronic obstructive pulmonary disease (AECOPD) frequency and severity is unknown. Method: We measured immunological biomarkers and clinical characteristics in 271 stable chronic obstructive pulmonary disease (COPD) patients (67% male, mean age 63 years) from "The Obstructive Pulmonary Disease Outcomes Cohort of Switzerland" cohort on a single occasion. One-year follow-up data were available for 178 patients. Variables independently associated with AECOPD frequency and severity were identified by multivariable regression analyses. Receiver operating characteristic analysis was used to obtain optimal cutoff levels and measure the area under the curve (AUC) in order to assess if baseline data can be used to predict future AECOPD. Results: Higher number of COPD medications (adjusted incident rate ratio [aIRR] 1.17) and platelet count (aIRR 1.03), and lower FEV1% predicted (aIRR 0.84) and IgG2 (aIRR 0.84) were independently associated with AECOPD frequency in the year before baseline. Optimal cutoff levels for experiencing frequent (>1) AECOPD were ≥3 COPD medications (AUC = 0.72), FEV1 ≤40% predicted (AUC = 0.72), and IgG2 ≤2.6 g/L (AUC = 0.64). The performance of a model using clinical and biomarker parameters to predict future, frequent AECOPD events in the same patients was fair (AUC = 0.78) but not superior to a model using only clinical parameters (AUC = 0.79). The IFN-lambda rs8099917GG-genotype was more prevalent in patients who had severe AECOPD. Conclusions: Clinical and biomarker parameters assessed at a single point in time correlated with the frequency of AECOPD events during the year before and the year after assessment. However, only clinical parameters had fair discriminatory power in identifying patients likely to experience frequent AECOPD. Keywords: Acute exacerbation of CO