This paper also produces the partial dependence plot
for atemp variable. After accounting for all the
variables, the marginal effect of the selected variables
is illustrated in Figure 8. As the paper presumed, the
count of bikes first increases with atemp and at about
0.6 it reaches the climax, then decreases with atemp.
Figure 8: Partial independence plot for atemp variable
(Picture credit: Original).
Figure 9 shows the fitted values and the actual values
of the data by the random forest model. The model
fitting effect is good, as the two lines in the
comparison chart are very close. This can be
illustrated through that the model’s fitted value is
much around to the actual value.
Figure 9: Fitting plot for random forest (Picture credit:
Original)
.
7 CONCLUSION
The study mainly focuses on the bike renting count
prediction using the Bike Rental Sharing Data Set.
The analyzing outcome shows find that random forest
model improves the predicting result and its mean
square value,
and cross validation error best
compared to backward stepwise selection regression,
ridge regression, lasso,bagging, and boosting. This
leads to the conclusion that the random forest model
can be seen as a useful tool to predict bike demand.
From the result variance importance analysis, it can
infer that the most influential factors of bike rental
demand are temperature, humidity, and wind speed.
This conclusion can contribute to bike system
operation to predict the bike using demand more
accurately and determine the redistribution of bikes
more precisely.
REFERENCES
J. C. García-Palomares, J. Gutiérrez, and M. Latorre,
“Optimizing the location of stations in bike-sharing
programs: A GIS approach,” Applied Geography, vol.
35, Jul.2012, pp. 235-246, doi:10.1016/j.apgeog.2012.
07.002.
J.Schuijbroek, R. C. Hampshire and W. J. Van Hoeve,
“Inventory rebalancing and vehicle routing in bike
sharing systems,” European Journal of Operational
Research, vol. 257, Jun. 2016, pp. 992 – 1004,
doi:10.1016/j.ejor.2016.08.029.
Z. Zheng, Y. Zhou, L. Sun, “A Multiple Factor Bike Usage
Prediction Model in Bike-Sharing System,” In: Li, S.
(eds) Green, Pervasive, and Cloud Computing GPC
2018. Lecture Notes in Computer Science, vol 11204,
2019.
C. Wirtgen, M. Kowald, J. Luderschmidt, H. Hünemohr,
“Multivariate demand forecasting for rental bike
systems based on an unobserved component model,”
Electronics, vol. 11, Dec. 2022, pp. 4146,
2022,doi:10.3390/electronics11244146.
C.Thirumalai, and R. Koppuravuri, “Bike sharing
prediction using deep neural networks,” JOIV:
International Journal on Informatics Visualization, vol.
1, 2017, pp. 83-87,doi:10.30630/joiv.1.3.30.
V. A. Alencar, L. R.Pessamilio, F. Rooke, H. S. Bernardino,
and A. Borges Vieira, “Forecasting the carsharing
service demand using uni and multivariable models,”
Journal of Internet Services and Applications, vol. 12,
Dec. 2021, pp. 1-20,doi:10.1186/s13174-021-00137-8.
S. Ruffieux, N. Spycher, E. Mugellini and O. A. Khaled,
"Real-time usage forecasting for bike-sharing systems:
A study on random forest and convolutional neural
network applicability," 2017 Intelligent Systems
Conference (IntelliSys), London, UK, 2017, pp. 622-
631, doi: 10.1109/IntelliSys.2017.8324359.
R. Salih-Elamin, and H. Al-Deek, “Short-term prediction
for bike share systems' travel time under the effects of
weather conditions,” Advances in transportation
studies, vol. 50. Apr. 2020, pp. 81-94, doi:10.4399/
97888255317326.
S. Li, “A Decision Model for Bike Sharing Based on Big
Data Analysis,” Journal Européen des Systèmes
Automatisés, vol. 53, Apr. 2020,pp 283-288,
doi:10.18280/jesa.530216.
H. Fanaee-T, and J. Gama, “Event labeling combining
ensemble detectors and background knowledge,”
Progress in Artificial Intelligence, vol. 2, Jun. 2014, pp.
113-127,doi:10.1007/s13748-013-0040-3.
S. VE, and Y Cho, Y. “A rule-based model for Seoul Bike
sharing demand prediction using weather data,”
European Journal of Remote Sensing, vol. 53, Jun2020,
pp. 166-183, doi:10.1080/22797254.2020.1725789.
G. James, D. Witten, T. Hastie and R. Tibshirani. An
introduction to statistical learning. New York: Springer,
vol. 112, 2011, pp. 18.