越來越多人想要自己投資理財產品,而不是將錢儲蓄在銀行裡。最近,參與投資俱樂部成為投資者的不錯選擇。跟隨投資俱樂部的財務專家與他們學習如何選取股票並獲得更高的投資報酬率是有利可圖。但許多投資俱樂部在2007-2008年的金融海嘯中表現不理想且其獲利變得低落。在這項研究中,我們希望提出一種推薦機制是結合群眾智慧和一種稱為長短期記憶(LSTM)的特殊類型的循環神經網絡,以推薦符合投資俱樂部需求的投資組合。根據俱樂部的風險承受能力和投資風格,我們的機制可以為俱樂部的投資者推薦適當的股票投資組合。利用StockTwits和股票歷史數據,驗證我們提議的投資組合推薦機制比其他市場基準表現更好。
More and more people want to invest financial products by themselves rather than saving in the banks. Recently, participating in an investment club becomes a good choice for investors. It is likely more profitable to follow with financial experts in investing club as they can learn how to select stock and get greater positive ROI. But many of investment clubs have low return of rate from financial crisis in 2007-2008. In this study, we want to propose a recommendation mechanism that combine collective intelligence and a special type of recurrent neural network called LSTM to recommend portfolio to accord with investment club demand. According to club's risk tolerance and investment style, our system can recommend appropriate stock portfolio for investors in the club. Utilizing StockTwits and stock historical data, we verify that the proposed portfolio recommendation mechanism performs better than other benchmarks in market.