The measured errors were categorically divided
into 2 groups, which are false negative and false positive.
According to the data acquired from 2011, 2010
and 2006. The former group is of higher proportion
than the later, i.e., 14.25, 15 and 18.5 percents of false
positive compared to 8.3, 6.6 and 0.9 percents of false
negative. In order to elucidate these findings, each of
the factors and their weights were analyzed. Similar
to the previous study [4], each factor was rated with
low, medium or high flood risk according Table 2.
It could then be deduced that, the false negative instance
was predicted because it contained less low and medium risk factors than the high risk ones. Those
lower risk factors are, based on 2011 data, of significant
neural weights. In this year, for instance, the
low - medium risk factors are annually and monthly
rainfall, flood plain in the past, height and sub-basin
are-as. Similarly, the false positive instance occurred
because it has more data of high risk factors than
those lower ones.
The measured errors were categorically dividedinto 2 groups, which are false negative and false positive.According to the data acquired from 2011, 2010and 2006. The former group is of higher proportionthan the later, i.e., 14.25, 15 and 18.5 percents of falsepositive compared to 8.3, 6.6 and 0.9 percents of falsenegative. In order to elucidate these findings, each ofthe factors and their weights were analyzed. Similarto the previous study [4], each factor was rated withlow, medium or high flood risk according Table 2.It could then be deduced that, the false negative instancewas predicted because it contained less low and medium risk factors than the high risk ones. Thoselower risk factors are, based on 2011 data, of significantneural weights. In this year, for instance, thelow - medium risk factors are annually and monthlyrainfall, flood plain in the past, height and sub-basinare-as. Similarly, the false positive instance occurredbecause it has more data of high risk factors thanthose lower ones.
การแปล กรุณารอสักครู่..
