초록

용어보기논문에서 용어와 풀이말을 자동 추출한 결과로, 시범 서비스 중입니다. AI-Helper

Abstract ▼ AI-Helper

This paper proposes a method for predicting the probability of fire occurrence through weather conditions in a region. The proposed prediction model is generated using a decision tree algorithm from a data set, which combines fire investigation data with weather data observed every hour. The fire in...

This paper proposes a method for predicting the probability of fire occurrence through weather conditions in a region. The proposed prediction model is generated using a decision tree algorithm from a data set, which combines fire investigation data with weather data observed every hour. The fire investigation data are irregularly generated, unlike the periodically generated weather data. A fire occurs at a specific location in a region, but a weather datum is measured over the region. We refer to a factor to be considered in calculating fire probability from a decision tree which is generated from the combined data set. We generated 16 decision trees, which predicts the probability of fire occurrence in the corresponding region, from fire investigation data and weather data generated during five years in each of seven major cities and nine states of South Korea. All of the 16 decision trees produced lower mean absolution errors than decision trees each of which consists of only the root node. They also produced an average of 111.37 rules in the form of IF-THEN statements. The rule set shows to be consistent with the assumption that fire and humidity have a close relationship.

주제어