Are mortgage loan closing delay risks predictable? A predictive analysis using text mining on discussion threads

Goldberg, David M.; Zaman, Nohel; Brahma, Arin; Aloiso, Mariano

doi:10.1002/asi.24559

Cited by 5 publications

(3 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Performance metrics for the study included mean absolute error (MAE), root mean square error (RMSE), and mean absolute percentage error (MAPE) to validate the forecasting performance of our proposed model. These three measures have been widely used in previous studies to examine the difference between actual and forecasted demand (Dai et al, 2022;Goldberg et al, 2022;Mukta et al, 2019). They are calculated as follows:…”

Section: Experiments Design and Performance Metricsmentioning

confidence: 99%

“…Performance metrics for the study included mean absolute error (MAE), root mean square error (RMSE), and mean absolute percentage error (MAPE) to validate the forecasting performance of our proposed model. These three measures have been widely used in previous studies to examine the difference between actual and forecasted demand (Dai et al, 2022; Goldberg et al, 2022; Mukta et al, 2019). They are calculated as follows:

MAE goodbreak= \frac{1}{n} \sum_{i = 1}^{n} (||, y_{i} goodbreak- \overset{y_{i}}{true})

RMSE goodbreak= \sqrt{\frac{1}{n} \sum_{i = 1}^{n} {(y_{i} - \overset{true¯}{y_{i}})}^{2}}

MAPE goodbreak= \frac{1}{n} \sum_{i = 1}^{n} \frac{|y_{i} - \overset{true¯}{y_{i}}|}{y_{i}}

where

y_{i}

is the observed hotel demand in week i , and

\overset{y_{i}}{true}

denotes the forecasted hotel demand in week i .…”

Section: Empirical Studymentioning

confidence: 99%

See 1 more Smart Citation

What online review features really matter? An explainable deep learning approach for hotel demand forecasting

Zhang

2023

Asso for Info Science & Tech

View full text Add to dashboard Cite

Accurate demand forecasting plays a critical role in hotel revenue management. Online reviews have emerged as a viable information source for hotel demand forecasting. However, existing hotel demand forecasting studies leverage only sentiment information from online reviews, leading to capturing insufficient information. Furthermore, prevailing hotel demand forecasting methods either lack explainability or fail to capture local correlations within sequences. In this study, we (1) propose a comprehensive framework consisting of four components: expertise, sentiment, popularity, and novelty (ESPN framework), to investigate the impact of online reviews on hotel demand forecasting; (2) propose a novel dual attention-based long short-term memory convolutional neural network (DA-LSTM-CNN) model to optimize the model effectiveness. We collected online review data from Ctrip.com to evaluate our proposed ESPN framework and DA-LSTM-CNN model. The empirical results show that incorporating features derived from the ESPN improves forecasting accuracy and our DA-LSTM-CNN significantly outperforms the state-of-the-art models. Further, we use a case study to illustrate the explainability of the DA-LSTM-CNN, which could guide future setups for hotel demand forecasting systems. We discuss how stakeholders can benefit from our proposed ESPN framework and DA-LSTM-CNN model.

show abstract

Section: Experiments Design and Performance Metricsmentioning

confidence: 99%

“…Performance metrics for the study included mean absolute error (MAE), root mean square error (RMSE), and mean absolute percentage error (MAPE) to validate the forecasting performance of our proposed model. These three measures have been widely used in previous studies to examine the difference between actual and forecasted demand (Dai et al, 2022; Goldberg et al, 2022; Mukta et al, 2019). They are calculated as follows:

MAE goodbreak= \frac{1}{n} \sum_{i = 1}^{n} (||, y_{i} goodbreak- \overset{y_{i}}{true})

RMSE goodbreak= \sqrt{\frac{1}{n} \sum_{i = 1}^{n} {(y_{i} - \overset{true¯}{y_{i}})}^{2}}

MAPE goodbreak= \frac{1}{n} \sum_{i = 1}^{n} \frac{|y_{i} - \overset{true¯}{y_{i}}|}{y_{i}}

where

y_{i}

is the observed hotel demand in week i , and

\overset{y_{i}}{true}

denotes the forecasted hotel demand in week i .…”

Section: Empirical Studymentioning

confidence: 99%

What online review features really matter? An explainable deep learning approach for hotel demand forecasting

Zhang

2023

Asso for Info Science & Tech

View full text Add to dashboard Cite

show abstract

“…[13] and Goldberg et al [14] each used smoke terms for 13 this purpose, and both found that follow-on machine learning models 14 approached the accuracy of deep learning word embedding models. 15…”

Section: Uses Of Smoke Terms For Text Mining 64mentioning

confidence: 99%