Machine Learning and Network Analytics in Finance
Permanent URI for this collection
1 - 3 of 3
ItemYou are What You Say: The Influence of Company Tweets on Its Stock Performance( 2019-01-08)This paper investigates the relationship between Twitter metrics and stock price performance of a company. The objective of this research is to contribute to the area of research that seeks to uncover the business value of social media platforms. Building on prior research, this paper identifies two categories of metrics that have been used to examine the relationship between Twitter metrics and stock performance of a company, namely traffic and motivation. While traffic is measured as volume of tweets, motivation is measured from two perspectives; polarity (positive, neutral, and negative) and emotion (positive emotion and negative emotion). Unstructured data from Twitter and Yahoo finance Website about Amazon was gathered to test the study hypothesis. A combination of machine learning techniques for text analytics and hierarchical regression analysis was used to analyze the data. Results indicate that emotional motivation expressed in tweets sent out by a company positively influences the company’s stock performance.
ItemA Tensor-based eLSTM Model to Predict Stock Price Using Financial News( 2019-01-08)Stock market prediction has attracted much attention from both academia and business. Both traditional finance and behavioral finance believe that market information affects stock movements. Typically, market information consists of fundamentals and news information. To study how information shapes stock markets, common strategies are to concatenate various information into one compound vector. However, such concatenating ignores the interlinks between fundamentals and news information. In addition, the fundamental data are continuous values sampled at fixed time intervals, while news information occurred randomly. Such heterogeneity leads to miss valuable information partially or twist the feature spaces. In this article, we propose a tensor-based event-LSTM (eLSTM) to solve these two challenges. In particular, we model the market information space with tensors instead of concatenated vectors and balance the heterogeneity of different data types with event-driven mechanism in LSTM. Experiments performed on an entire year data of China Securities markets demonstrate the supreme of the proposed approach over the state-of-the-art algorithms including AZfinText, eMAQT, and TeSIA.