Page 74 - Contributed Paper Session (CPS) - Volume 7

CPS2031 Javier Linkolk L. et al.
Finally, according to Figure 5,6 and 7, the LSTM are presented, where each
square represents a neuron, that is, each neuron reflects an LSTM. This shows
that there are 25 LSTMs that generate the 365 time series.
In parallel, in figure 8, the root of the mean quadratic error (RMSE) of each
neuron is observed along with its Train Score (TSN), reporting the lowest RMSE

in the neuron (1,3).

4. Discussion and Conclusion
Under the LSTM model, the estimation and adjustment report of the "La
Florida" station had important results close to the real data. In the same way
that [19] in this research it is shown that the LSTM has a good capacity of
adaptation to estimate the concentration of PM2.5, even more when it merges
with the SOM at the time of grouping.
The assembly between SOM and LSTM, allowed grouping the time series
according to the determined pattern. This showed neurons that have a group
of similar time series that corresponds to the pollution record per hour in a
day.
For each neuron an LSTM network was adjusted, able to model its data
well; thus, one LSTM is trained per neuron, but each LSTM learns several time
series.
As future work, different assembled algorithms will be competed, both for
estimation and clustering, with the assembly proposal that was made.

References
1. Liang, B., Li, X. L., Ma, K., and Liang, S. X, “Pollution characteristics of
metal pollutants in PM2. 5 and comparison of risk on human health in
heating and non-heating seasons in Baoding, China”, Ecotoxicology and
environmental safety, vol. 170, pp. 166-171, 2019.
2. Xing, Y. F., Xu, Y. H., Shi, M. H., and Lian, Y. X, “The impact of PM2. 5 on
the human respiratory system”, Journal of thoracic disease, vol. 8, no 1,
p. E69, 2016.
3. Wang, C., Tu, Y., Yu, Z., & Lu, R, “PM2. 5 and cardiovascular diseases in
the elderly: An overview”, International journal of environmental
research and public health, vol. 12, no 7, p. 8187¬8197, 2015.
4. Zhang, Q., Jiang, X., Tong, D., Davis, S. J., Zhao, H., Geng, G., and Ni, R,
“Transboundary health impacts of transported global air pollution and
international trade”, Nature, vol. 543, no 7647, p. 705, 2017.
5. Atkinson, R. W., Kang, S., Anderson, H. R., Mills, I. C., and Walton, H. A.,
“Epidemiological time series studies of PM2. 5 and daily mortality and
hospital admissions: a systematic review and meta-analysis”, Thorax,
2014, p. thoraxjnl-2013-204492, 2014.

61 | I S I W S C 2 0 1 9