Herein, we have compared the performance of SVM and MLP in emotion recognition using speech and song channels of the RAVDESS dataset. We have undertaken a journey to extract various audio features, identify optimal scaling strategy and hyperparameter for our models. To increase sample size, we have performed audio data augmentation and addressed data imbalance using SMOTE. Our data indicate that optimised SVM outperforms MLP with an accuracy of 82 compared to 75%. Following data augmentation, the performance of both algorithms was identical at ~79%, however, overfitting was evident for the SVM. Our final exploration indicated that the performance of both SVM and MLP were similar in which both resulted in lower accuracy for the speech channel compared to the song channel. Our findings suggest that both SVM and MLP are powerful classifiers for emotion recognition in a vocal-dependent manner.
The COVID-19 pandemic has caused ~ 2 million fatalities. Significant progress has been made in advancing our understanding of the disease process, one of the unanswered questions, however, is the anomaly in the case/mortality ratio with Mexico as a clear example. Herein, this anomaly is explored by spatial analysis and whether mortality varies locally according to local factors. To address this, hexagonal cartogram maps (hexbin) used to spatially map COVID-19 mortality and visualise association with patient-level data on demographics and pre-existing health conditions. This was further interrogated at local Mexico City level by choropleth mapping. Our data show that the use of hexagonal cartograms is a better approach for spatial mapping of COVID-19 data in Mexico as it addresses bias in area size and population. We report sex/age-related spatial relationship with mortality amongst the Mexican states and a trend between health conditions and mortality at the state level. Within Mexico City, there is a clear south, north divide with higher mortality in the northern municipalities. Deceased patients in these northern municipalities have the highest pre-existing health conditions. Taken together, this study provides an improved presentation of COVID-19 mapping in Mexico and demonstrates spatial divergence of the mortality in Mexico.
Disadvantaged groups around the world have suffered and endured higher mortality during the current COVID-19 pandemic. This contrast disparity suggests that socioeconomic and health-related factors may drive inequality in disease outcome. To identify these factors correlated with COVID-19 outcome, country aggregate data provided by the Lancet COVID-19 Commission subjected to correlation analysis. Socioeconomic and health-related variables were used to predict mortality in the top 5 most affected countries using ridge regression and extreme gradient boosting (XGBoost) models. Our data reveal that predictors related to demographics and social disadvantage correlate with COVID-19 mortality per million and that XGBoost performed better than ridge regression. Taken together, our findings suggest that the health consequence of the current pandemic is not just confined to indiscriminate impact of a viral infection but that these preventable effects are amplified based on pre-existing health and socioeconomic inequalities.