資訊技術絕對是企業取得競爭優勢功不可沒的角色。不僅是將人工轉為機器自動化、 將紙本文件轉為電子數據等數位化(Digitalization)動作，而是運用資訊技術進行整合管理、 數據分析、決策輔助等提升商業價值的行為。更因如此，近期多數企業在思考公司未來 成長發展時，都開始將數位轉型(Digital Transformation) 考慮在內。 在眾多的數位技術中，近幾年機器學習(Machine Learning)發展蓬勃，在各領域皆有 很好的成效，如客製化行銷、醫療診斷輔助、語音及圖形識別。引領了許多企業紛紛投 入資源，無論是與資料科學專家合作抑或自行訓練資訊專業人才。但是該如何成功導入 機器學習至企業當中，企業所面臨的困難與挑戰是什麼?特別是以中小企業在資源與資 金使用的限制條件下，他們該如何有效的準備及運用? 本研究以如何運用機器學習並實作數據分析，進而改進製程良率為目標。採用台灣 某製造高爾夫球桿頭公司的生產數據資料，搭配決策樹演算法(Decision Tree)提出對應 之訓練模型與評估方式，並根據生產數據之訓練結果建立規則，進而輔助企業調整生產 過程的製程參數及控制生產過程的變因。;Information technology plays the critical role in business success. Companies gain competitive advantage by adopt IT strategies. Not only automation for manpower reduction, paper to digital transformation. Companies use information technology to integrated business management, data analytics and decision supporting for increase business value. Therefore, digital transformation strategy becomes the most important plan for many companies. These days, machine learning is the most popular and important technology. It is evolving at such a rapid pace, enabling great results in many areas. Such as marketing personalization, clinical decision support, voice and image recognition. Machine learning adoption is increasing in many industries. But, what are the difficulties and challenges firms face? Especially, the small and medium-sized enterprises which with the resource limitation. The research focus on build the application of data analytics on yield rate. Use the golf club head manufacturing process data which provided by an anonymous company. The research builds the training model and evaluation method with decision tree algorithm. According to the training result, provides the suggestion which help the firm to adjust manufacturing parameters and avoid producing defective products.
