Research Area: big data, computational intelligence, knowledge graph, deep learning
Speech Title: Forecasting daily stock trend using multi-filter feature selection and deep learning
Abstract：Stock market forecasting has attracted significant attention mainly due to the potential monetary benefits. Predicting these markets is a challenging task due to numerous interrelated factors, and needs a complete and efficient feature selection process to identify the most informative factors. As a time series problem, stock price movements are also dependent on movements on its previous trading days. Feature selection techniques have been widely applied in stock forecasting, but existing approaches usually use a single feature selection technique, which may overlook some important assumptions about the underlying regression function linking the input and output variables. In this study, we combine features selected by multiple feature selection techniques to generate an optimal feature subset and then use a deep generative model to predict future price movements. First, we compute an extended set of forty-four technical indicators from daily stock data of eighty-eight stocks and then compute their importance by independently training logistic regression model, support vector machine and random forests. Based on a prespecified threshold, the lowest ranked features are dropped and the rest are grouped into clusters. The variable importance measure is reused to select the most important feature from each cluster to generate the final subset. The input is then fed to a deep generative model comprising of a market signal extractor and an attention mechanism. The market signal extractor recurrently decodes market movement from the latent variables to deal with stochastic nature of the stock data and the attention mechanism discriminates between predictive dependencies of different temporal auxiliary outputs. The results demonstrate that combining features selected by multiple feature selection approaches and using them as input into a deep generative model outperforms state-of-the-art approaches.
A.Prof. Heng Liu
Sun Yat-sen University
Research Area: 1. Innovation, entrepreneurship and business model 2. Corporate strategy and management 3. Management research methods (empirical research, meta-analysis) 4. Institutions, culture and Chinese management issues
Speech Title: The paradox of management tone and managerial risk-taking: Evidence from textual sentiment analysis
Abstract：Despite textual sentiment analysis has become a rising technique in organizational research with the exponential growth of computer power and the availability of textual data, existing literature offers quite limited understanding on how management tone influences corporate decision makings. Integrated from the lens of prospect theory and expectation violation theory, we examine how management tone affects a firm’s risky behavior. The study finds that managers tend to take more risks after they release either more positive or negative tones in earning communication conferences, resulting in an overall U-shaped relationship between management tone and corporate risk taking. The study also finds that such role of tone upon risk-taking tend to be even stronger when the firm has higher reputation. Taken together, these findings contribute to expectation violation theory and corporate textual sentiment literatures by suggesting that managers’ framing of stakeholder expectation is a double-edged sword and strongly influences their risky decision-making.
A. Prof. Yanjun Qian
Northwestern Polytechnical University
Research Area：Marketing Research; Production; Business Administration; Project Management Optimization; Product Development
Speech Title: Information Management in New Product Development
Abstract：Facing intense competition, rapidly evolving technologies, changing customer needs, and shorter product life cycles, many firms need to develop lower cost, higher quality products at a rapid pace. Efficient new product development is essential to achieve these goals, and thus is critical to the success of many modern corporations. However, structuring new product development is challenging. Part of the difficulty is due to the complex information relationship among design activities. Unlike the manufacturing process, new product development often involves a number of decision-making activities, for example, the design of an automobile may involve thousands of engineers making millions of design decisions. Moreover, none of these activities are performed in isolation; instead, each design choice may affect many other design parameters.
Fortunately, the Design Structure Matrix (DSM) provides a concise way in describing and investigating information dependencies among design activities. A DSM is a matrix representation of a project with elements denoting individual activities which are executed in the temporal order listed from top to bottom. Sub-diagonal marks represent information input from upstream activities to downstream, and super-diagonal marks denote feedbacks from downstream activities to upstream. This study will introduce how to apply the DSM method for efficient new product development. Using a case of balancing machine development, I will first introduce how to establish the information dependencies among the activities and build a DSM. Then, I will introduce how to reorganize the interrelated activities in a DSM so as to structure new product development.