scholarly journals Using Scaling Methods to Improve Support Vector Regression’s Performance for Travel Time and Traffic Volume Predictions

Author(s):  
Amanda Yan Lin ◽  
Mengcheng Zhang ◽  
Selpi
2020 ◽  
Author(s):  
Lewis Mervin ◽  
Avid M. Afzal ◽  
Ola Engkvist ◽  
Andreas Bender

In the context of bioactivity prediction, the question of how to calibrate a score produced by a machine learning method into reliable probability of binding to a protein target is not yet satisfactorily addressed. In this study, we compared the performance of three such methods, namely Platt Scaling, Isotonic Regression and Venn-ABERS in calibrating prediction scores for ligand-target prediction comprising the Naïve Bayes, Support Vector Machines and Random Forest algorithms with bioactivity data available at AstraZeneca (40 million data points (compound-target pairs) across 2112 targets). Performance was assessed using Stratified Shuffle Split (SSS) and Leave 20% of Scaffolds Out (L20SO) validation.


2015 ◽  
Vol 2015 ◽  
pp. 1-9 ◽  
Author(s):  
Cong Bai ◽  
Zhong-Ren Peng ◽  
Qing-Chang Lu ◽  
Jian Sun

Accurate and real-time travel time information for buses can help passengers better plan their trips and minimize waiting times. A dynamic travel time prediction model for buses addressing the cases on road with multiple bus routes is proposed in this paper, based on support vector machines (SVMs) and Kalman filtering-based algorithm. In the proposed model, the well-trained SVM model predicts the baseline bus travel times from the historical bus trip data; the Kalman filtering-based dynamic algorithm can adjust bus travel times with the latest bus operation information and the estimated baseline travel times. The performance of the proposed dynamic model is validated with the real-world data on road with multiple bus routes in Shenzhen, China. The results show that the proposed dynamic model is feasible and applicable for bus travel time prediction and has the best prediction performance among all the five models proposed in the study in terms of prediction accuracy on road with multiple bus routes.


2013 ◽  
Vol 25 (5) ◽  
pp. 445-455 ◽  
Author(s):  
Fang Zong ◽  
Jia Hongfei ◽  
Pan Xiang ◽  
Wu Yang

This paper presents a model system to predict the time allocation in commuters’ daily activity-travel pattern. The departure time and the arrival time are estimated with Ordered Probit model and Support Vector Regression is introduced for travel time and activity duration prediction. Applied in a real-world time allocation prediction experiment, the model system shows a satisfactory level of prediction accuracy. This study provides useful insights into commuters’ activity-travel time allocation decision by identifying the important influences, and the results are readily applied to a wide range of transportation practice, such as travel information system, by providing reliable forecast for variations in travel demand over time. By introducing the Support Vector Regression, it also makes a methodological contribution in enhancing prediction accuracy of travel time and activity duration prediction.


Author(s):  
Long Tien Truong ◽  
Majid Sarvi ◽  
Graham Currie

Numerous studies have explored design and evaluation of bus lane priority by using empirical, analytical, and simulation approaches. However, none attempted to understand how different bus lane combinations, such as continuous and discontinuous bus lane sections, and a different number of bus lane sections, affect bus performance and general traffic. This paper investigates operational effects of bus lane combinations to establish whether multiple bus lane sections create a multiplier effect in which a series of continuous bus lane sections creates more benefits than several single-lane sections. If a multiplier effect exists, it suggests scale economies in wider implementation of bus priority on a networkwide scale. Overall, results confirm that there is a multiplier effect; thus bus travel time benefits and general traffic travel time disbenefits are proportional to the number of links with a bus lane. The effect suggests a constant return to scale on continuous multiple sections. The results also suggest that converting a traffic lane to a bus lane when the upstream traffic volume exceeds the capacity of the remaining traffic lanes causes significant negative effects for buses and general traffic. In addition, negative general traffic effects of continuous bus lane combinations are lower than those for a similar number of discontinuous bus lanes. Bus delays at intersections approaching the bus lane tend to improve when upstream traffic volume does not exceed the capacity of remaining downstream traffic lanes. Policy implications and areas for future research are suggested.


2014 ◽  
Vol 587-589 ◽  
pp. 2230-2233
Author(s):  
Qian Nan Jiao ◽  
Jian Jun Wang ◽  
Teng Fei Zhang

This paper use the different layout forms of bus lanes as the study subject, useing the micro-simulation tool VISSIM to comparative analysis different layout forms’ passing traffic volume and travel time in the different traffic volume、traffic flow rate conditions. And offer related suggestions of respective forms’ adaptability.


2020 ◽  
Vol 5 (3) ◽  
pp. 275-281
Author(s):  
Onyemaechi John Nnamani ◽  
Victor Ayodele Ijaware ◽  
Joseph Olalekan Olusina ◽  
Timothy Oluwadare Idowu

Travel time variability or distribution is very important to travel time reliability studies in transportation systems. This study aimed at developing a multivariate regression model for estimating travel times for dynamic highway networks in Akure Metropolis. The independent variables for the model are Traffic volume, density, speed of vehicles, and traffic flow while the dependent response variable is the Travel time. The estimated travel time was compared with the observed travel time from the real field data and the estimation using the regression model reveals a significant level of accuracy. Also, it was discovered that traffic volume, speed, density, and flow were highly correlated with travel time. The result analyzed using descriptive statistics in the SPSS software environment reveals an R2 value of 0.998, thereby indicating that the independent variables accounted for 99% of travel time in the study area. The Hypothesis tested at 95% confidence level using ANOVA unveils that there is no significant difference between the observed and estimated travel time model. The Mean Absolute Percentage Error (MAPE) of 0.049 shows that the model performed very well and was very efficient for analyzing the probabilistic relation between travel time and the independent variables. The study recommends the use of the developed travel time model for estimating travel time within the study area.


Sign in / Sign up

Export Citation Format

Share Document