Forex trading is one of the most volatile markets which basically depend on currency exchange. There are so many situations like disasters, political activities, etc. which affects the exchange rate of currency. Prediction of Price or Exchange rate can be done either via technical approach or fundamental approach, but there are some flaw exists in both the approaches, so we propose a sentiment scored based algorithmic approach which is a hybrid model that overcomes the shortcomings of existing approaches. The experimental results show that the proposed method gives the better accuracy as compare to traditional approaches.

 

Published In : IJCSN Journal Volume 4, Issue 4

Date of Publication : August 2015

Pages : 643 - 649

Figures :03

Tables : 08

Publication Link : Sentiment Score based Algorithmic Trading

 

 

 

Sukesh Kumar Ranjan : Department of Computer Science and Engineering / Information Technology, Jaypee Institute of Information Technology, Sector-62 Noida, Uttar Pradesh, India

Abhishek Trivedi : Department of Computer Science and Engineering / Information Technology, Jaypee Institute of Information Technology, Sector-62 Noida, Uttar Pradesh, India

Dharmveer Singh Rajpoot : Department of Computer Science and Engineering / Information Technology, Jaypee Institute of Information Technology, Sector-62 Noida, Uttar Pradesh, India

 

 

 

 

 

 

 

Trading

Sentiment

Profit

Loss

Risk

We have find the accuracy of 69.32 which is higher to the other benchmarks this signifies that our proposed model is predicting the value precisely than other existing model. Along with all these, there are some silent features of our model 1. Our model is using very less data for computation. 2. We are considering all news for the prediction rather than depending on single news for prediction. 3. Our combined approach of fundamental trading and technical trading is also removing the major drawback of technical trading as well as fundamental trading. Along with features we have also some limitations. In the bubble situation, our model is not allowing to trade. But in future we will try overcome this limitation.

 

 

 

 

 

 

 

 

 

[1] Xiaodong Li, Haoran Xie,Ran Wang,Yi Cai,Jingjing Cao,Feng Wang,Huaqing Min,Xiaotie Deng, Empirical analysis: stock market prediction via extreme learning machine, London: Springer-Verlag, 2014. [2] Arne Thorvald Gierløff Hollum, Borre P. Mosch, and Zoltán Szlávik, Economic Sentiment: Text-Based Prediction of Stock Price Movements with Machine Learning and WordNet , Berlin Heidelberg : Springer- Verlag,2013. [3]http://thomsonreuters.com/content/dam/openweb/documents/ pdf/tr-com-financial/news-analytics-product-brochure-- oct-2010.pdf. [4] Currency Composition of Official Foreign Exchange Reserves (COFER), International Financial Statistics (IFS),” http://data.imf.org/”. [5] http://www.dukascopy.com [6] EleftheriosSoulas, Dennis Shasha “Online Machine Learning AlgorithmsFor Currency Exchange Prediction” TR-2013-953 [7] Megan Potoski” Predicting Gold Prices” CS229,2013 [8] Chenxu Shao, Zheming Zheng “Algorithmic trading using machine learning techniques”2013 [9] Gabriel Pui Cheong Fung, Jeffrey Xu Yu, and Wai Lam “News Sensitive Stock Trend Prediction”. [10] Jinli Hu, Amos Storkey “Multi-period Trading Prediction Markets with Connections to MachineLearing.