Research on carpooling demand prediction study based on machine learning

doi:10.3969/j.issn.1674-8484.2024.05.010

Abstract

Abstract:

In order to improve the accuracy of carpooling demand prediction, thereby enhancing the efficiency of ride hailing services and effectively alleviating traffic congestion, a regional carpooling probability prediction model was proposed by optimizing the traditional decision tree machine learning model using time feature extraction and Kepler optimization algorithm. An experiment was conducted to predict carpooling demand based on the Chicago ride hailing probability dataset, and the model was compared with traditional decision tree models. The experimental results show that the optimized model outperforms traditional decision tree models in terms of prediction accuracy, with a decrease of 0.044 in mean absolute error (MAE) and 0.054 in root mean squared error (RMSE). The optimized model has higher accuracy in predicting carpooling demand compared to traditional decision tree models.

Key words: shared mobility, carpooling demand, machine learning, decision tree, Kepler optimization algorithm

CLC Number:

U121

WANG Di, LI Ying, HU Yujiao, SUN Haocheng. Research on carpooling demand prediction study based on machine learning[J]. Journal of Automotive Safety and Energy, 2024, 15(5): 723-731.

Figures/Tables 17

References 18

[1]	黄雪原. 一种改进 kNN 算法及在即时拼车需求预测中的应用[D]. 大连: 大连理工大学, 2018.
	HUANG Xueyuan. An improved kNN algorithm and its application in real-time carpooling demand prediction[D]. Dalian: Dalian University of Technology, 2018. (in Chinese)
[2]	Altshuler T, Altshuler Y, Katoshevski R, et al. Modeling and prediction of ride-sharing utilization dynamics[J]. J Advan Transport, 2019, 2019(1): No 6125798
[3]	龚学辉. 基于位置预测的共乘群体发现与优化方法研究[D]. 西安: 长安大学, 2019.
	GONG Xuehui. Research on the discovery and optimization method of shared ride groups based on location prediction[D]. Xi'an: Chang'an University, 2019. (in Chinese)
[4]	Piao, Yong. Ad Hoc Ridesplitting Prediction Based on Improved kNN[C]// 2019 12th Int'l Symp Comput Intel Desig (ISCID). Hangzhou, China,IEEE, 2019, 2: 162-167.
[5]	Fuloria S. Passenger demand forecasting in the ridesharing context: A comparison of statistical and deep learning approaches[J]. IUP J Appl Econom, 2020, 19(1) : 57-71.
[6]	LI Xing, ZHANG Ying, DU Ming, et al. The forecasting of passenger demand under hybrid ridesharing service modes: A combined model based on WT-FCBF-LSTM[J]. Sustain Citi Soc, 2020, 62: 102419.
[7]	毛林冉. 基于出行需求预测的动态拼车匹配算法及最小车队规划方法研究[D]. 西安: 西北大学, 2021.
	MAO Linran. Research on dynamic carpooling matching algorithm and minimum fleet planning method based on travel demand prediction[D]. Xi'an: Northwestern University, 2021. (in Chinese)
[8]	杜明洋. 城市网约车拼车时空特征分析与派单优化研究[D]. 南京: 东南大学, 2022.
	DU Mingyang. Research on spatiotemporal characteristics analysis and dispatching optimization of urban ride hailing carpooling[D]. Nanjing: Southeast University, 2022. (in Chinese)
[9]	DU Juan, Rakha H A, Breuer H. An in-depth spatio- temporal analysis of ride-hailing travel: The Chicago case study[J]. Case Stud Transport Poli, 2022, 10(1): 118-129.
[10]	芝加哥市政府. 芝加哥市政府区域开放数据[DB/OL]. (2019-03-01). https://www.Wunderground.com.
	City of Chicago. Chicago city government area open data[DB/OL]. (2019-03-01). https://www.Wunderground.com.
[11]	Weather Underground. 芝加哥市2019年天气[DB/OL]. (2019-03-01). https://www.wunderground.com.
	Weather Underground. Chicago weather in 2019[DB/OL]. (2019-03-01). https://www.wunderground.com.
[12]	Hastie T, Tibshirani R, Friedman J, et al. Random forests[J]. Elem Statist Learn: Data Mini, Infer, Predict, 2009(2): 587-604.
[13]	CHEN Tianqi, Guestrin C. Xgboost: A scalable tree boosting system[C]// Proc 22nd Acm sigKDD Int’l Conf Knowl Disco Data Mining. San Francisco, CA, USA,2016: 785-794.
[14]	KE Guolin, MENG Qiwei, Finley T, et al. Lightgbm: A highly efficient gradient boosting decision tree[J]. Advan Neur Info Proce Syst, 2017, 30: 3146-3154.
[15]	Prokhorenkova L, Gusev G, Vorobev A, et al. CatBoost: unbiased boosting with categorical features[J]. Advan Neur Info Proce Syst, 2018, 31: 6638-6648.
[16]	Rainforth T, Le T A, van de Meent J W, et al. Bayesian optimization for probabilistic programs[J]. Advan Neur Info Proce Syst, 2016, 29: 280-288.
[17]	黄德山. 基于 BERT 模型的国内股票评论情感分析[D]. 北京: 北京大学, 2023.
	HUANG Deshan. Sentiment analysis of domestic stock reviews based on BERT model[D]. Beijing: Peking University, 2023. (in Chinese)
[18]	Greenacre M, Groenen P J F, Hastie T, et al. Principal component analysis[J]. Nat Rev Meth Prim, 2022, 2(1): 100.

字段	字段说明	示例数据
AreaNo	区域编号	1
Name	名称	Rogers Park
Crowded%	拥挤占比/ %	7.7
16+Unemployed%	16岁及以上失业人口占比/ %	8.7
25+NoHighSchoolDiploma%	25岁以上无高中文凭人口占比/ %	18.2
Under18Over64%	18岁以下64岁以上人口占比/ %	27.5
IncomeMean	人均收入/ $	23939
HardshipIndex	贫困指数/ %	39

字段	字段说明	示例数据
AreaNo	区域编号	1
Name	名称	Rogers Park
Crowded%	拥挤占比/ %	7.7
16+Unemployed%	16岁及以上失业人口占比/ %	8.7
25+NoHighSchoolDiploma%	25岁以上无高中文凭人口占比/ %	18.2
Under18Over64%	18岁以下64岁以上人口占比/ %	27.5
IncomeMean	人均收入/ $	23939
HardshipIndex	贫困指数/ %	39

字段	字段说明	示例数据
Area	上车的区域	77
Date	时间段起始时间	2019/1/26 14∶15
Trip_Seconds_Median	行驶时间中位数/ s	705
Trip_Miles_Median	行驶里程中位数	2.95
Fare_Median	订单价格中位数	6.25
Shared_Trip_Authorized_Sum	拼车订单数	5
Total_Count	总订单数	20
Shared_Trip_Authorized_Ratio	拼车率	0.25
Temperature(°F)	温度/ (°F)	2.2
Bad_weather	是否下雨或下雪	1
Humidity / %	湿度/ %	89

字段	字段说明	示例数据
Area	上车的区域	77
Date	时间段起始时间	2019/1/26 14∶15
Trip_Seconds_Median	行驶时间中位数/ s	705
Trip_Miles_Median	行驶里程中位数	2.95
Fare_Median	订单价格中位数	6.25
Shared_Trip_Authorized_Sum	拼车订单数	5
Total_Count	总订单数	20
Shared_Trip_Authorized_Ratio	拼车率	0.25
Temperature(°F)	温度/ (°F)	2.2
Bad_weather	是否下雨或下雪	1
Humidity / %	湿度/ %	89

模型	MAE	RMSE
随机森林	0.107 12	0.154 58
XGBoost	0.084 38	0.101 23
LightGBM	0.085 47	0.110 68
CatBoost	0.075 46	0.098 95