以需求方平台觀點預測在即時競標系統中線上廣告流量之方法

Abstract

隨著線上廣告產業興起，對需求方平台來說如何掌握廣告流量以精確的控制預算花 費變成一個重要的議題。然而需求方平台與供應方平台不同的是，他們難以拿到即時 的觀看者以及網頁、手機應用程式的資訊，就算拿到了也必須以非常短的時間決策、 回應供應方平台的廣告需求，大量的特徵會拖垮我們的預測速度。有鑒於此，我們在 本論文提出一個從需求方平台的角度預測廣告流量的方法。我們使用更精簡、更容易 取得的特徵，以及有閉型解的迴歸模型加速我們的流量預測。除此之外，我們的方法 能辨別流量異常並予以處理，也能跟上長期的趨勢。我們最後大約1億7千萬筆測試資 料中預測總誤差約0.9%，平均每單位時間(本篇以小時為單位)誤差大約11%。

Online advertising has been all the rage these years. Budget control and traffic prediction turn out to be important issues for the demand-side platforms(DSP). However, DSPs cannot easily grab the information of audiences and media platforms. Although DSPs might have the information immediately, it is still hard to response the request of advertisements in realtime due to the high volume of features. Therefore, we propose a method predicting traffic of requests of advertisements from perspective of DSPs. The features we used are more simple and easy to be extracted from history data. The prediction model we chose is regression model with closed-form solution. Both the features and regression model make our prediction adaptive in real-time systems. Our method can detect traffic anomalies and prevent it from overwhelming prediction. Moreover, our method can also keep pace of the trend. Experiment results show that our method’s error rate of prediction is about 0.9% in total, and 10% per time unit.