Search

In this paper we study neural networks and their approximating power in panel data models. We provide asymptotic guarantees on deep feed-forward neural network estimation of the conditional mean, building on the work of Farrell et al. (2021), and explore latent patterns in the cross-section. We use the proposed estimators to forecast the progression of new COVID-19 cases across the G7 countries during the pandemic. We find significant forecasting gains over both linear panel and nonlinear time-series models. Containment or lockdown policies, as instigated at the national level by governments, are found to have out-of-sample predictive power for new COVID-19 cases. We illustrate how the use of partial derivatives can help open the "black box" of neural networks and facilitate semi-structural analysis: school and workplace closures are found to have been effective policies at restricting the progression of the pandemic across the G7 countries. But our methods illustrate significant heterogeneity and time variation in the effectiveness of specific containment policies. [ABSTRACT FROM AUTHOR]

Neural Networks (NNs) are essential components of intelligent systems because they produce efficient solutions to problems of overwhelming complexity for conventional computing methods. There are lots of papers showing that NNs can approximate a wide variety of functions, but comparatively very few discuss their limitations. To illustrate the role played by information coding in NN computations we define and study sensitive and robust Boolean functions. We also prove that an exponential large set of functions in the first group are exponentially difficult to compute by multiple-layer perceptrons (hence incomputable by single-layer perceptrons) and a comparatively large set of functions in the second one, but not all, are computable by single-layer perceptrons. This suggests that the success of NNs, or lack of it, are in part determined by properties of the learned data sets. Finally we use polynomial threshold perceptrons to compute all Boolean functions with quantum annealing and present in detail a QUBO computation on the D-Wave Advantage. [ABSTRACT FROM AUTHOR]

This paper uses a dataset of more than 900,000 news stories to test whether news can predict stock returns. We measure sentiment with a proprietary Thomson-Reuters neural network. We find that daily news predicts stock returns for only 1 to 2 days, confirming previous research. Weekly news, however, predicts stock returns for one quarter. Positive news stories increase stock returns quickly, but negative stories have a longdelayed reaction. Much of the delayed response to news occurs around the subsequent earnings announcement. [ABSTRACT FROM AUTHOR]

We present an indicator of job loss derived from Twitter data, based on a fine-tuned neural network with transfer learning to classify if a tweet is job-loss related or not. We show that our Twitter-based measure of job loss is well-correlated with and predictive of other measures of unemployment available in the official statistics and with the added benefits of real-time availability and daily frequency. These findings are especially strong for the period of the Pandemic Recession, when our Twitter indicator continues to track job loss well but where other real-time measures like unemployment insurance claims provided an imperfect signal of job loss. Additionally, we find that our Twitter job loss indicator provides incremental information in predicting official unemployment flows in a given month beyond what weekly unemployment insurance claims offer. [ABSTRACT FROM AUTHOR]

Recent modeling developments have created tradeoffs between attribution-based models, models that rely on causal relationships, and "pure prediction models" such as neural networks. While forecasters have historically favored one technology or the other based on comfort or loyalty to a particular paradigm, in domains with many observations and predictors such as textual analysis, the tradeoffs between attribution and prediction have become too large to ignore. We document these tradeoffs in the context of relabeling 27 million Thomson Reuters news articles published between 1996 and 2021 as debt-related or non-debt related. Articles in our dataset were labeled by journalists at the time of publication, but these labels may be inconsistent as labeling standards and the relation between text and label has changed over time. We propose a method for identifying and correcting inconsistent labeling that combines attribution and pure prediction methods and is applicable to any domain with human-labeled data. Implementing our proposed labeling solution returns a debt-related news dataset with 54% more observations than if the original journalist labels had been used and 31% more observation than if our solution had been implemented using attribution-based methods only. [ABSTRACT FROM AUTHOR]

In recent times, AI technologies have made huge waves in the healthcare industry, sparking a heated debate over whether AI doctors could ultimately replace human doctors. Machines will not replace human doctors in the near future, but AI can help physicians make better medical judgments or even substitute human judgment in certain sectors such as healthcare. The recent successful applications of AI in healthcare have been enabled by the growing usage of healthcare data and the fast growth of big data analytic approaches. Effective AI strategies, driven by relevant clinical questions, can uncover clinically relevant knowledge contained in vast quantities of data, assisting clinical decision-making. In light of this, this research was carried out to shed light on studies and research that discussed the use of artificial intelligence in the fields of healthcare and disease diagnostics. in addition to providing the appropriate background to the research community in order to advance this area, and to assist in the integration of a variety of AI technologies around patient needs in a variety of healthcare settings, especially for chronic care patients with the most complex comorbidities and care requirements. The results of the research confirm that AI encompasses a variety of areas of the system, including thinking, information representation, searching for solutions, and machine learning (ML), to name a few. ML is a form of machine learning that uses algorithms to recognize patterns in a set of data that has been further classified. Deep learning (DL) is a subfield of machine learning that uses Artificial Neural Networks (ANNs). In addition, Artificial intelligence can help in developing drug formulation, aid in decision-making, diagnosing diseases, and evaluating the best treatment for a patient, such as customized drugs, dealing with the medical information produced and using it for potential drug production and AI is also expected to play an important role in the production of pharmaceutical products. [ABSTRACT FROM AUTHOR]

The article explores seasonal adjustment methods for consumer price indexes (CPI). It discusses the use of an X-12/X-13 method and standard econometric tests by national statistical offices (NSO). It examines the use of machine learning, particularly neural networks, in improving seasonal adjustment of CPI.

A list of research studies by Air Force Institute of Technology's (AFIT) students that are sponsored by the Air Combat Command of the U.S. Air Force is presented which includes "Metric Development for Continuous Process Improvement," "Strategic Implications of US Fighter Force Reductions: Air-to-Air Combat Modeling Using Lanchester Equations" and "Using Upper Layer Weights to Efficiently Construct and Train Feedforward Neural Networks Executing Backpropagation."

Labyrinth weirs provide an economic option for flow control structures in a variety of applications, including as spillways at dams. The cycles of labyrinth weirs are typically placed in a linear configuration. However, numerous projects place labyrinth cycles along an arc to take advantage of reservoir conditions and dam alignment, and to reduce construction costs such as narrowing the spillway chute. Practitioners must optimize more than 10 geometric variables when developing a head--discharge relationship. This is typically done using the following tools: empirical relationships, numerical modeling, and physical modeling. This study applied a new tool, machine learning, to the analysis of the geometrically complex arced labyrinth weirs. In this work, both neural networks (NN) and random forests (RF) were employed to estimate the discharge coefficient for this specific type of weir with the results of physical modeling experiments used for training. Machine learning results are critiqued in terms of accuracy, robustness, interpolation, applicability, and new insights into the hydraulic performance of arced labyrinth weirs. Results demonstrate that NN and RF algorithms can be used as a unique expression for curve fitting, although neural networks outperformed random forest when interpolating among the tested geometries. [ABSTRACT FROM AUTHOR]

In this work we investigate the possibility to use state of the art approaches for deep learning on point clouds for matching segments, sensed from objects under varying perspectives, with the aim to obtain positional information for relative positioning. Therefore we propose a method and evaluate it using a custom real world data set. [ABSTRACT FROM AUTHOR]