Page 30 - Contributed Paper Session (CPS) - Volume 3

CPS1934 Atikur R. K. et al.

Figure 1. Data integration procedure

In this paper, we only consider daily episodes of respiratory tract infection
(RTI) that includes both lower and upper respiratory tract infection, daily
minimum and maximum temperatures, wind speed, sea level pressure, and
relative humidity. Both MySQL quires and R routines have been used to
process, analyse and visualize our data. We construct rolling time series
statistics from these climatic variables to predict RTI episodes.
Let us compute rolling statistics for some climatic variables. If () is an

instance of a climatic time series of the th weather station at time (day) , then
−lagged rolling mean and standard deviation can be computed as

1
̅ (|) = ∑ =1 ( − + ) (1)

1
2
= √ ∑ ( ( − + ) − ̅ (|) ) (2)
=1

Let () and () are maximum and minimum temperatures at time
. The difference between maximum and minimum temperatures, () =
() − (), indicates the amount of fluctuation in a day. Assuming that
() is the difference between the maximum and minimum temperatures of

weather station on day , we compute −day rolling mean (|) and
̅

standard deviation (|) by using Eq.(1) and Eq.(2). Similarly, for the th

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