“…This pattern is consistent with an analysis of the individual routes in different cities (see Supplementary Note 11 for additional scatter and contour density plots of the distribution of routes in ξ - D -space) and confirmed by an analysis of the rank statistics of across cities, strongly suggesting that the median of the detour heterogeneity is systematically lower for informal services than it is for formal ones (see Table 1 and Supplementary Note 13) . Strong detour heterogeneity and convex detour profiles pose additional inconvenience at the beginning and end of the routes, often to commuters traveling to and from the outskirts of a city, adding to their already long travel times due to the distance covered and affecting them more than those traveling closer to city centers 38 – 41 . Fig.…”
The Global South, encompassing more than 80% of the world population, heavily relies on informal paratransit services with ad-hoc routes. Yet, it remains unclear how efficiently such informal public transport services organize and operate. Here, we analyze and compare the structural efficiency of more than 7000 formal and informal bus service routes in 36 cities across 22 countries globally. Intriguingly, informal transport self-organizes in ways at or above efficiency levels of centralized services. They exhibit fewer detours, more uniform paths, and comparable interconnectivities, all while remaining profitable without the major subsidies common in the Global North. These insights challenge the global perception of informal transport as an inferior alternative to centrally organized services. More generally, analyzing large-scale microscopic transport data and condensing them into informative macroscopic observables may qualitatively improve system understanding and reveal specific options to create more accessible, efficient, and sustainable public transport solutions.
“…This pattern is consistent with an analysis of the individual routes in different cities (see Supplementary Note 11 for additional scatter and contour density plots of the distribution of routes in ξ - D -space) and confirmed by an analysis of the rank statistics of across cities, strongly suggesting that the median of the detour heterogeneity is systematically lower for informal services than it is for formal ones (see Table 1 and Supplementary Note 13) . Strong detour heterogeneity and convex detour profiles pose additional inconvenience at the beginning and end of the routes, often to commuters traveling to and from the outskirts of a city, adding to their already long travel times due to the distance covered and affecting them more than those traveling closer to city centers 38 – 41 . Fig.…”
The Global South, encompassing more than 80% of the world population, heavily relies on informal paratransit services with ad-hoc routes. Yet, it remains unclear how efficiently such informal public transport services organize and operate. Here, we analyze and compare the structural efficiency of more than 7000 formal and informal bus service routes in 36 cities across 22 countries globally. Intriguingly, informal transport self-organizes in ways at or above efficiency levels of centralized services. They exhibit fewer detours, more uniform paths, and comparable interconnectivities, all while remaining profitable without the major subsidies common in the Global North. These insights challenge the global perception of informal transport as an inferior alternative to centrally organized services. More generally, analyzing large-scale microscopic transport data and condensing them into informative macroscopic observables may qualitatively improve system understanding and reveal specific options to create more accessible, efficient, and sustainable public transport solutions.
“…Since the patterns of non-linear relationships vary among built environment variables, this makes parametric specification of non-linear relationships inefficient and inaccurate. Apart from that, it is suggested that applying the threshold of built environment variables during the urban planning process would help achieve the goal of a low carbon city ( 51 ).…”
Understanding the relationship between CO2 emissions from commuting (CEC) and the built environment is crucial for sustainable transportation and land-use policymaking during the process of constructing a low carbon city. Previous studies usually assume that the relationship is linear, which may lead to inaccurate CEC prediction and ineffective policy. Using daily travel survey data of residents in the central city of Jinan, this study adopted a gradient boosting decision tree model to explore the threshold effect and the non-linear relationship between built environments and CEC. Our findings suggest that 40% of CEC is related to the workplace environment, which is higher than the residential environment and other socioeconomic variables. The five most important variables are road density within 1 km radius of the workplace (13.493%), distance to the center at workplace and residence (10.908%, 10.530%), population density at workplace (9.097%) and distance to bus stop from the residence (8.399%). Distance to city center plays the most important role and its non-linear relationship reflects the influence of the urban spatial structure of Jinan on CEC. Furthermore, the thresholds and non-linear relationships provide planning guidelines to support urban planning development policies for low carbon city.
“…One way to visualize “SHAP importance values” (or SHAP importances) is with a beeswarm plot. 37 Each dot in the plot represents a single data point (in our case, a hospital admission in the NIS). The y axis shows the different preoperative factors, and the x axis shows the SHAP importance for that factor.…”
BACKGROUND:
A range of statistical approaches have been used to help predict outcomes associated with colectomy. The multifactorial nature of complications suggest that machine learning algorithms may be more accurate in determining postoperative outcomes by detecting nonlinear associations, which are not readily measured by traditional statistics.
OBJECTIVE:
The aim of this study was to investigate the utility of machine learning algorithms to predict complications in patients undergoing colectomy for colonic neoplasia.
DESIGN:
Retrospective analysis using decision tree, random forest, and artificial neural network classifiers to predict postoperative outcomes.
SETTINGS:
National Inpatient Sample database (2003-2017).
PATIENTS:
Adult patients who underwent elective colectomy with anastomosis for neoplasia.
INTERVENTIONS(S) IF ANY:
N/A.
MAIN OUTCOME MEASURES:
Performance was quantified using sensitivity, specificity, accuracy, and area-under-the-curve-receiver-operator-characteristic to predict the incidence of anastomotic leak, prolonged length of stay, and inpatient mortality.
RESULTS:
A total of 14,935 patients (4,731 laparoscopic, 10,204 open) were included. They had an average age of 67±12.2 years and 53% were female. The three machine learning models successfully identified patients who developed the measured complications. Although differences between model performance were largely insignificant, the neural network scored highest for most outcomes: predicting anastomotic leak, area-under-the-curve-receiver-operator-characteristic 0.88/0.93 (open/laparoscopic, 95% CI, 0.73-0.92/0.80-0.96); prolonged length of stay, area-under-the-curve-receiver-operator-characteristic 0.84/0.88 (open/laparoscopic, 95% CI, 0.82-0.85/0.85-0.91); and inpatient mortality, area-under-the-curve-receiver-operator-characteristic 0.90/0.92 (open/laparoscopic, 95% CI, 0.85-0.96/0.86-0.98).
LIMITATIONS:
The patients from the National Inpatient Sample database may not be an accurate sample of the population of all patients undergoing colectomy for colonic neoplasia and does not account for specific institutional and patient factors.
CONCLUSIONS:
Machine learning predicted postoperative complications in patients with colonic neoplasia undergoing colectomy with good performance. Though validation using external data and optimization of data quality will be required, these machine learning tools show great promise in assisting surgeons with risk-stratification of perioperative care to improve postoperative outcomes. See Video Abstract.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
