Results

Found 3 articles

Clear filters

Read the News, Not the Books: Forecasting Firms’ Long-term Financial Performance via Deep Text Mining
PDF
2023 | ACM Transactions on Management Information Systems | Citations: 0
Authors: Zhai, Shuang (Sophie); Zhang, Zhu (Drew)

Abstract: In this paper, we show textual data from firm-related events in news articles ca ... Expand

Abstract: In this paper, we show textual data from firm-related events in news articles can effectively predict various firm financial ratios, with or without historical financial ratios. We exploit state-of-the-art neural architectures, including pseudo-event embeddings, Long Short-Term Memory Networks, and attention mechanisms. Our news-powered deep learning models are shown to outperform standard econometric models operating on precise accounting historical data. We also observe forecasting quality improvement when integrating textual and numerical data streams. In addition, we provide in-depth case studies for model explainability and transparency. Our forecasting models, model attention maps, and firm embeddings benefit various stakeholders with quality predictions and explainable insights. Our proposed models can be applied both when numerically historical data is or is not available. Collapse

Semantic filters: Box-Jenkins method

Topics: missing data accounting financial performance sensemaking organizational productivity

Methods: deep learning autoregressive integrated moving average longitudinal research time series analysis case study
Forecasting medical cost inflation rates: A model comparison approach
PDF
2012 | Decision Support Systems | Citations: 3
Authors: Cao, Qing; Ewing, Bradley T.; Thompson, Mark A.

Abstract: Due to healthcare costs rising faster than overall cost of living, decision make ... Expand

Abstract: Due to healthcare costs rising faster than overall cost of living, decision makers (i.e., households, businesses, and governments) must cut back on healthcare utilization or spending elsewhere to be fiscally responsible. Accurate forecasts of future medical costs are critical for efficient planning, budgeting and operating decisions at all levels. This research compares the accuracy of the linear autoregressive moving average (ARMA) model and the nonlinear neural network model in producing forecasts of medical cost inflation rates. The analysis focuses on twelve monthly measures of medical costs including the overall medical care price index and eleven (disaggregated) subsectors of medical costs. In addition to standard symmetric measures of forecast accuracy, we utilize two asymmetric error measures designed to capture and penalize preferences for under- and overprediction in model selection. The findings indicate that the neural network model outperforms the univariate ARMA in both 1-step and 12-step ahead forecasts. A number of important practical implications are discussed, such as the use of accurate forecasts in contract negotiations, budgeting and planning. Collapse

Semantic filters: Box-Jenkins method

Topics: decision support system logistics management artificial intelligence

Methods: autoregressive moving average model artificial neural network longitudinal research time series analysis paired t-test
A two-stage neural network approach for ARMA model identification with ESACF
PDF
1994 | Decision Support Systems | Citations: 0
Authors: Lee, Jae Kyu; Jhee, Won Chul

Abstract: We attempt to design artificial neural netwirks that can help in the automatic i ... Expand

Abstract: We attempt to design artificial neural netwirks that can help in the automatic identification of the Autoregressive Moving Average (ARMA) model. For this purpose, we adopt the Extended Sample Autocorrelation Function (ESACF) as a feature extractor, and the Multi-Layered Percepttron as a Pattern Classification Network. Since the performance test from the network is sensitive to the noise in input ESACF patterns, we suggest a preprocessing Noise Filtering Network. It turns out that the Noise Filtering Network significantly improves the performance. To reduce the computational burden of training the full Pattern Classification Network, we suggest a Reduced Network that can still perform as good as the full network. The two-stage filtering and classifying networks performed very well (90% of accuracy) not only with the artificially generated data sets but also with the real world time series. We have also reconfirmed that the performance of ESACF is superior to that of ACF and PACF. Collapse

Semantic filters: Box-Jenkins method

Topics: expert system

Methods: autoregressive moving average model longitudinal research design artifact time series analysis computational algorithm

Page 1 of 1

AIS11

Journals

Conferences

Semantic Filter

Found 3 articles

Read the News, Not the Books: Forecasting Firms’ Long-term Financial Performance via Deep Text Mining

Forecasting medical cost inflation rates: A model comparison approach

A two-stage neural network approach for ARMA model identification with ESACF