隨著物價上漲、薪資縮減、女性意識抬頭，現代人晚婚、不婚已成普遍的現象，養育一名小孩所費不貲，綜合以上原因，婦女生育率逐年下滑，這也導致眾多國家面臨了少子化的社會問題，正因為生育率低下，如何降低新生兒死亡率便是一項重要任務，因此，本研究希望透過個案醫院產科所提供的資料集，結合機器學習的技術來找出對於新生兒健康發展存在重要影響的特徵。 本研究共設計了兩階段的實驗，第一階段是去探討新生兒出生前的特徵對於新生兒出生當下健康狀態的影響，第二階段則是去探討新生兒出生前以及住院期間的特徵對於新生兒能否健康出院的影響，所有實驗皆使用五種分類器來建構預測模型，最後希望實驗結果能供醫護人員在新生兒護理決策上做參考，降低新生兒因健康狀態不佳而導致不良結局的情況，達到精準健康促進的目標。;It has become a common phenomenon for modern people to marry late and not to marry resulting from price hike, wages stagnation, and the awakening female consciousness. Moreover, it costs to raise a child. Due to many reasons, the fertility rate of women has been declining year by year and this has also caused many social problems. With the low fertility rate, it is more important to control the newborns’ mortality rate. Therefore, we hope to use machine learning techniques with the maternal and newborn datasets provided by the hospitals to find out the significant variables for newborns’ health development. We designed a two-stage experiment. The first stage is to explore the impact of the newborns’ characteristics before birth on the health of the newborns. The second stage is to explore the characteristics of the newborn before birth and during the hospital stay for the impact of whether the newborn can be discharged healthily from the hospital. All experiments use five classifiers to construct the predictive models. In the end, we hope that the results of our study can be used as a reference for medical staff to make decisions on newborn’s care and reduce the adverse outcomes of the newborn, achieving the goal of precision health promotion.
