如何制定與驗證有效的投資策略一直是金融科技領域持續討論的焦點之一。在制定策略之前，從眾多因素中挑選出合適的因子是一項繁瑣的工程。若投資者能夠先對資料進行有系統的綜覽，了解資料的潛在趨勢，將能更有效率地制定投資策略。 為了達到這個目的，本研究提出了一套探索性資料分析（EDA）平台。透過濾網進行初步篩選，我們可以縮小股數及因子討論範圍。並利用熱力圖、散布回歸圖、盒鬚圖、Facegrid組圖等統計圖，將股價常見相關因子的趨勢以視覺化方式呈現。這將有助於經驗不足的投資者選擇因子的方向，也能協助經驗豐富的投資者優化策略。為了符合投資者的自身投資屬性，我們保留了客製化與系統化的空間，讓投資者可以透過XML進行自定義分析因子、設定參數分析圖表及其瀏覽流程，打造專屬的投資策略研究方法。 本研究透過三個實驗模擬在基本面、技術面、籌碼面的濾網下，展示了如何使用本系統獲得潛力因子的視覺化資訊。使用者可以將本工具與其他回測平台搭配使用，最終建立出最合適的投資策略。;How to formulate an effective investment strategy has always been a focal point in the financial technology field. Before devising a strategy, selecting suitable factors from numerous considerations is a meticulous task. If investors can systematically overview the data and understand its underlying trends, they can formulate investment strategies more efficiently. To achieve this goal, this study proposes an Exploratory Data Analysis (EDA) platform. Through a filtering tool for preliminary screening, we can narrow down the scope of discussions. The trends of stock-related factors are presented visually by means of statistical charts, such as heatmaps, scatter plots, box plots, Facegrid composite plots, etc. This helps inexperienced investors in choosing the direction of factors and assists experienced investors in optimizing their strategies. To cater to investors′ personalized investment preferences, we provide customization options. Investors can customize analysis factors, set analysis parameters, analyze charts, and use XML to navigate the process, creating a personalized investment strategy research approach. This study demonstrates how to obtain visual information on potential factors through the following three aspects: fundamental, technical, and chip. Users can integrate this tool with other backtesting platforms to ultimately establish the most suitable investment strategy.
