|
on Big Data |
| By: | Shuozhe Li; Zachery B Schulwol; Risto Miikkulainen |
| Abstract: | To the naked eye, stock prices are considered chaotic, dynamic, and unpredictable. Indeed, it is one of the most difficult forecasting tasks that hundreds of millions of retail traders and professional traders around the world try to do every second even before the market opens. With recent advances in the development of machine learning and the amount of data the market generated over years, applying machine learning techniques such as deep learning neural networks is unavoidable. In this work, we modeled the task as a multivariate forecasting problem, instead of a naive autoregression problem. The multivariate analysis is done using the attention mechanism via applying a mutated version of the Transformer, "Stockformer", which we created. |
| Date: | 2025–01 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2502.09625 |
| By: | Haavio, Markus; Heikkinen, Joni; Jalasjoki, Pirkka; Kilponen, Juha; Paloviita, Maritta |
| Abstract: | We study the evolution of the European Central Bank's (ECB) monetary policy since July 2021, following the adoption of a new strategy and amid a period of volatile inflation. Utilizing text analysis, we assess changes in the general sentiment of the ECB's communication. Additionally, we employ topic modeling to develop an inflation focused tone index. By integrating these tone indices with real-time data from monetary policy meetings, we directly estimate the ECB's loss function. Our findings indicate a recent shift towards a more inflation-centered communication approach by the ECB. Preliminary results also suggest that the ECB's policy preferences have become more symmetric since July 2021. |
| Keywords: | asymmetric loss function, central bank communication, textual analysis, topic model, optimal monetary policy |
| JEL: | E31 E52 E58 |
| Date: | 2025 |
| URL: | https://d.repec.org/n?u=RePEc:zbw:bofrdp:313643 |
| By: | Yue Qiu (Finance School, Shanghai University of International Business and Economics, Shanghai, China); Wenbin Wang (Finance School, Shanghai University of International Business and Economics, Shanghai, China); Tian Xie (College of Business, Shanghai University of Finance and Economics, Shanghai, China); Jun Yu (Faculty of Business Administration, University of Macau, Macao); Xinyu Zhang (Academy of Mathematics and Systems Science, Chinese Academy of Sciences, Beijing, China) |
| Abstract: | Many real-world analytics problems, such as forecasting sales of fashion products, involve uncertain and heterogeneous demand, requiring prescriptive analytics to incorporate multiple covariates and address the inherent challenge of model uncertainty. Traditional predict-thenoptimize (PTO) approaches typically rely on a single predictive model, overlooking model uncertainty. To address this, we propose an ensemble learning framework that integrates Mallows-type model averaging into the PTO paradigm, leveraging diverse candidate models with varying covariates to enhance forecast accuracy and decision robustness. Theoretically, we prove that the weighted forecasts achieve asymptotic optimality under mild conditions and establish finite-sample risk bounds, ensuring stable performance even in limited-data settings. We empirically evaluate the proposed framework using weekly store-level sales data from an internationally recognized footwear brand in China. The forecasting exercise demonstrates that our approach consistently achieves the lowest prediction risk, improving forecast accuracy by 4.72% to 7.41% compared to the best-performing alternatives without weighted forecast features. In the subsequent decision optimization exercise, we identify gift, combo, and discount promotions as key decision variables and show that our framework delivers the highest predicted sales responses on average, outperforming alternative forecasting methods and existing data-driven decision frameworks. |
| Keywords: | data-driven, model uncertainty, model averaging, prescriptive analytics, machine learning, fashion sales forecasting |
| Date: | 2025–03 |
| URL: | https://d.repec.org/n?u=RePEc:boa:wpaper:202525 |
| By: | Louis Colnot (CSIL scrl) |
| Abstract: | CERN is advancing plans for the Future Circular Collider (FCC), a next-generation particle accelerator designed to push the boundaries of high-energy physics beyond what was possible using the Large Hadron Collider (LHC). As a scientific mega-project of unprecedented scale, the FCC requires substantial financial and scientific resources over a long time horizon, making sustained public support essential for its success. However, fostering favourable public perception is a complex challenge influenced by diverse factors, including the project’s scientific foundations, potential benefits, and the broader social and political context. Media outlets play a pivotal role in this process by selecting, framing, and broadcasting information to the public. This role persists despite the pressures exerted on traditional journalism by the rise of social media. This paper seeks to deepen our understanding of how large scientific projects are perceived in the media, using CERN and its LHC as a case study. Specifically, it examines how CERN and the LHC have been portrayed in French media from 2006 to 2024, focusing on the topics emphasised and the sentiment associated with this coverage. Two key research questions guide the analysis: (1) What are the main topics covered by French journalists regarding the LHC and CERN? (2) How has media perception of LHC and CERN evolved?. Using statistical and natural language processing techniques, I show that French media coverage of CERN and the LHC fluctuated cyclically, often influenced by major scientific milestones, such as the discovery of the Higgs boson, or negative incidents, such as terrorism allegations. Despite these fluctuations, French media coverage remained predominantly positive, showing no significant trend toward increased negativity or polarisation, as may be expected from comparisons with social media. These findings suggest that traditional French media may support fostering positive public perceptions of CERN and the LHC. The implications for research institutions’ communication strategies are also discussed. For instance, the positive perceptions of CERN/LHC in French media constitute a favourable context for the transition towards new projects, such as the FCC |
| Keywords: | Media, articles, perception, science, society, journalism, framing, topic modelling, sentiment analysis, LHC, CERN, FCC |
| JEL: | L82 I23 |
| Date: | 2025–03–04 |
| URL: | https://d.repec.org/n?u=RePEc:mst:wpaper:202501 |