Diabetes mellitus is a major noncommunicable disease, and reliable early-detection prediction can support Sustainable Development Goal (SDG) Target 3.4. This study compares standard logistic regression, Firth's Penalized Logistic Regression, and SMOTEN-based logistic regression for diabetes risk prediction under severe class imbalance using Wave 5 of the Indonesia Family Life Survey (IFLS5). The analysis included 115 respondents, comprising 8 diabetes cases (6.96%), with all predictors represented as categorical variables. Model performance was evaluated using 50 repeated stratified 75/25 train-test splits, with SMOTEN applied only to training data to prevent leakage. Standard logistic regression failed to converge across all partitions. Firth's Penalized Logistic Regression identified waist circumference as significantly associated with diabetes status (odds ratio = 10.69, 95% CI: 1.5673.14) and achieved mean accuracy of 89.17%, sensitivity of 8.00%, specificity of 95.19%, and AUC of 0.692. SMOTEN-based logistic regression achieved accuracy of 86.90%, sensitivity of 12.00%, specificity of 92.44%, and AUC of 0.632. Brier scores were 0.089 and 0.098, respectively, compared with 0.064 for the prevalence benchmark, indicating poorer overall probabilistic prediction performance. These findings illustrate the challenges of rare-event prediction under severe class imbalance and highlight the need for larger samples and further validation before clinical application.
Employee Stock Options (ESO) are a form of compensation given to employees in the form of shares that can be purchased within a specified period. The ESO valuation uses the Black Scholes model, assuming employees remain employed until the vesting period is completed, excluding the employee’s departure from the company before the option rights are exercised. This study aims to determine the value of Employee Stock Options (ESO) by adding a constant exit rate factor. The data from this study are the closing price at BANK XYZ for the calculation of Volatility, for the risk-free interest rate from the Federal Reserve Bank of St. Louis used to calculate the option value using the Black-Scholes model, adjusted for the probability of continued employment (survival probability) obtained from various constant exit rate values. The results of the study using the constant exit rate produce a lower Employee Stock Option value than the calculation using the standard Black-Scholes model. The decline is greater with increasing maturity time because the employee’s probability of remaining employed at the company decreases. By applying a constant exit rate, the assessment results better reflect the real conditions compared to the Black-Scholes model, which does not consider the risk of employee departure from the company
Stunting remains one of Indonesia’s most persistent public-health problems, and the nationalprevalence conceals subtantial variation across the country’s 38 provinces. This study modelsseven latent constructs behind stunting, antenatal care, early initiation of breastfeeding,immunization, sanitation, exclusive breastfeeding, low birth weight, and stunting using PartialLeast Squares Structural Equation Modeling on 2023–2024 provincial health-profile datapublished by Indonesia’s Ministry of Health, and then applies Finite Mixture Partial LeastSquares (FIMIX-PLS) to detect unobserved heterogeneity across provinces and to translate theresulting segments into a thematic map. The measurement model satisfied every recommendedvalidity and reliability threshold, with an Average Variance Extracted between 0.835 and0.932 and both Cronbach’s alpha and composite reliability above 0.80. In the pooled nationalmodel, low birth weight was the only construct with a statistically significant direct path tostunting (β = 0.540, p .001); antenatal care showed no significant direct path to stunting(β = −0.143, p = .678) but exerted a significant total effect (β = −0.365) entirely through afully mediated pathway via low birth weight (indirect effect β = −0.263, p = .022), togetherexplaining 63.0% of the variance in stunting. Model-selection criteria favored a three-segmentsolution comprising 21, 10, and 7 provinces, dominated respectively by sanitation-related,birth-weight-related, and breastfeeding-related structural pathways. Mapping these segmentsgeographically shows that provinces sharing a dominant pathway are not scattered at randombut tend to cluster by region, indicating that stunting-reduction programs should be tailoredto each segment’s dominant pathway rather than applied uniformly across the whole country.
The genus Panthera, comprising the lion, jaguar, leopard, tiger, and snow leopard, exhibitscomplex vocalization characteristics as biological communication mechanisms in its naturalenvironment. However, manual acoustic identification is limited by environmental variability,individual sound variations, and recording noise, necessitating an automated approach.This study applies the Higuchi Fractal Dimension (HFD) for feature extraction and theK-Nearest Neighbor (KNN) algorithm for classification to recognize vocalizations of fiveextant Panthera species. The dataset consists of 100 recordings (20 per species) from theTierstimmenarchiv repository, preprocessed by noise reduction, two-second segment selection,mono conversion, resampling, and amplitude normalization. An empirical fundamentalfrequency (F0) verification showed a range of 53 to 367 Hz, confirming the adequacy of theDWT decomposition depth. The signals were decomposed using DWT (Daubechies 4, 7levels), and HFD was computed on the resulting coefficients with Kmax of 10, 16, 28, 32, 50,and 60. Performance was evaluated using repeated nested stratified cross-validation (5-foldouter loop repeated 10 times; 5-fold inner loop for selecting Kmax and k), together with anablation study of DWT approximation (cA) and detail (cD) coefficients and a non-DWTbaseline computed directly from the preprocessed signal. The highest mean accuracy, 70.80%(SD = 8.77%), was obtained by combining all detail coefficients with the three highest-levelapproximation coefficients, and was stable across randomly sampled seeds (SD = 1.06%).The non-DWT baseline reached only 56.40%. These results indicate that HFD computedon DWT sub-band provides discriminative features for classifying Panthera vocalizations,although repeated individuals across folds cannot be ruled out.
Reporting dates do not identify claim payment times. We quantify a reporting-based payment-time assumption's effect on finite-horizon cash ruin in an idealized marine portfolio. The freMarine calibration comprises 1,109 positive-payment snapshot claims over 41 occurrence months, with negative-binomial monthly counts, finite-mixture severity, and empirical zero-inflated reporting lags. Refitted-bootstrap diagnostics assess distributional adequacy; conflicting severity diagnostics and rejected parametric lag fits remain explicit. Event-driven simulation couples immediate and proxy payments on the same occurrences and severities, with constant aggregate premiums and an empty initial system. For 250,000 paired paths over 60 months at initial surplus 2,000 EUR thousands, cash-ruin estimates are 0.2951/0.2363 for immediate/proxy payment. The paired difference is -0.0588 (95% conditional Monte Carlo interval [-0.0597,-0.0579]); mean outstanding payments are 718.79 EUR thousands. Nonnegative delay mechanically favors coupled cash surplus without discharging liabilities. Unobserved payment histories, opening claims, and policyholder behavior preclude claims of predictive superiority, a preferred handling policy, or regulatory capital needs.
