Keshab Raj Dahal
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Keshab Raj Dahal has a total of 21 co-authors. Here are the top co-authors by number of shared publications:
- Nawa Raj Pokhrel
Department of Physics and Computer Science, Xavier University of Louisiana, New Orleans, LA 70125, USA
3 shared publications Binod Rimal
2 shared publications- Binod Rimal
Department of Mathematics, The University of Tampa, Tampa, USA
2 shared publications Hum Nath Bhandari
2 shared publications- Hum Nath Bhandari
Department of Mathematics, Roger Williams University, Bristol, USA
2 shared publications Nawa Raj Pokhrel
2 shared publications
Latest Publications
- Open Access
- 18 June 2026
- Published by IgMin Publications Inc. in IgMin Research
- Vol. 4 (6) , 191-207
Deep Learning-Based Prediction of Nepal Stock Exchange Movements from Financial News Headlines
Predicting the direction of stock market movements is a challenging task due to its fuzzy, chaotic, volatile, nonlinear, and complex nature. However, with advancements in artificial intelligence, abundant data availability, and improved computational capabilities, creating robust models capable of accurately predicting stock market movement is now feasible. This study aims to develop a predictive model using news headlines to forecast the direction of stock market movements. It conducts a comparative analysis of four supervised classification deep learning models —long short-term memory (LSTM), gated recurrent unit (GRU), bidirectional long short-term memory (BiLSTM), and bidirectional gated recurrent unit (BiGRU)—to predict the next day’s movement direction of the close price of the Nepal Stock Exchange (NEPSE) index. Sentiment scores from the news headlines are computed using the Valence Aware Dictionary for Sentiment Reasoning (VADER) and the TextBlob sentiment analyzer. The models’ performance is evaluated based on sensitivity, specificity, accuracy, and the area under the receiver operating characteristic (ROC) curve (AUC). Experimental results indicate that all four models perform similarly when using sentiment scores from either VADER or TextBlob. Additionally, GRU and BiGRU models show consistent performance across both sentiment analyzers. However, LSTM and BiLSTM perform slightly better with TextBlob sentiment scores compared to those from VADER. These findings are further validated through statistical tests.
- 18 June 2026
- Published by IgMin Publications Inc. in IgMin Research
- Vol. 4 (6) , 191-207
Open AccessDeep Learning-Based Prediction of Nepal Stock Exchange Movements from Financial News Headlines
Predicting the direction of stock market movements is a challenging task due to its fuzzy, chaotic, volatile, nonlinear, and complex nature. However, with advancements in artificial intelligence, abundant data availability, and improved computational capabilities, creating robust models capable of accurately predicting stock market movement is now feasible. This study aims to develop a predictive model using news headlines to forecast the direction of stock market movements. It conducts a comparative analysis of four supervised classification deep learning models —long short-term memory (LSTM), gated recurrent unit (GRU), bidirectional long short-term memory (BiLSTM), and bidirectional gated recurrent unit (BiGRU)—to predict the next day’s movement direction of the close price of the Nepal Stock Exchange (NEPSE) index. Sentiment scores from the news headlines are computed using the Valence Aware Dictionary for Sentiment Reasoning (VADER) and the TextBlob sentiment analyzer. The models’ performance is evaluated based on sensitivity, specificity, accuracy, and the area under the receiver operating characteristic (ROC) curve (AUC). Experimental results indicate that all four models perform similarly when using sentiment scores from either VADER or TextBlob. Additionally, GRU and BiGRU models show consistent performance across both sentiment analyzers. However, LSTM and BiLSTM perform slightly better with TextBlob sentiment scores compared to those from VADER. These findings are further validated through statistical tests.
- Open Access
- 1 January 2026
- Published by Wiley in BioMed Research International
- Vol. 2026 (1) , e6298673
Food Away From Home and Body Mass Index: Insights From a Quantile Regression Analysis of NHANES 2011–2020
Background: Food away from home (FAFH) has several health implications and is believed to be a contributing factor to the rise in obesity. However, studies focusing on the impact of FAFH along the distribution of body mass index (BMI) remain scarce. This cross‐sectional study is aimed at using nationally representative survey data over multiple years to examine whether FAFH affects individuals differently based on their BMI. Methods: Data from the four recent cycles of the US National Health and Nutrition Examination Survey (NHANES) were pooled for analysis. Quantile regression (QR) method was utilized to model the outcome of interest “BMI” and the exposure “FAFH,” adjusted for potential confounders. The use of QR allowed differential effects of FAFH on BMI to be evident that would otherwise be masked using a linear or logistic regression model. Survey weights were applied in all analyses to address the complex sampling procedure of the NHANES. Results: QR method showed differential effects of FAFH on conditional BMI distribution. For individuals in the lower tail of the BMI distribution (0.05 quantile), three or more FAFH a week was associated with BMI increase of 0.29 kg/m 2 (95% CI: 0.00, 0.58) compared with less than three FAFH. The effect became stronger at the middle quantiles with the strongest effect observed at the 0.60 quantile, where individuals consuming three or more FAFH a week had a BMI of 0.41 kg/m 2 (95% CI: 0.12, 0.71) higher than those consuming FAFH less frequently. Although higher consumption of FAFH was associated with increased BMI for individuals in lower to middle quantiles, the effect at the upper tail (0.90 quantile), representing individuals in the highest BMI range, was not statistically significant. Additional investigation at the upper quantiles revealed that the association remained statistically significant at 0.70 quantile, whereas estimates at the 0.80 and 0.95 quantiles remained positive but were not statistically significant. Conclusion: QR revealed a broadly positive FAFH‐BMI association across the BMI distribution, with statistically significant associations through the 70th percentile and less precise estimates in the upper tail. Although further research is needed to better understand these associations, our findings highlight the importance of considering distributional patterns and statistical uncertainty when evaluating the impact of FAFH consumption on BMI and designing public health interventions.
- 1 January 2026
- Published by Wiley in BioMed Research International
- Vol. 2026 (1) , e6298673
Open AccessFood Away From Home and Body Mass Index: Insights From a Quantile Regression Analysis of NHANES 2011–2020
Background: Food away from home (FAFH) has several health implications and is believed to be a contributing factor to the rise in obesity. However, studies focusing on the impact of FAFH along the distribution of body mass index (BMI) remain scarce. This cross‐sectional study is aimed at using nationally representative survey data over multiple years to examine whether FAFH affects individuals differently based on their BMI. Methods: Data from the four recent cycles of the US National Health and Nutrition Examination Survey (NHANES) were pooled for analysis. Quantile regression (QR) method was utilized to model the outcome of interest “BMI” and the exposure “FAFH,” adjusted for potential confounders. The use of QR allowed differential effects of FAFH on BMI to be evident that would otherwise be masked using a linear or logistic regression model. Survey weights were applied in all analyses to address the complex sampling procedure of the NHANES. Results: QR method showed differential effects of FAFH on conditional BMI distribution. For individuals in the lower tail of the BMI distribution (0.05 quantile), three or more FAFH a week was associated with BMI increase of 0.29 kg/m 2 (95% CI: 0.00, 0.58) compared with less than three FAFH. The effect became stronger at the middle quantiles with the strongest effect observed at the 0.60 quantile, where individuals consuming three or more FAFH a week had a BMI of 0.41 kg/m 2 (95% CI: 0.12, 0.71) higher than those consuming FAFH less frequently. Although higher consumption of FAFH was associated with increased BMI for individuals in lower to middle quantiles, the effect at the upper tail (0.90 quantile), representing individuals in the highest BMI range, was not statistically significant. Additional investigation at the upper quantiles revealed that the association remained statistically significant at 0.70 quantile, whereas estimates at the 0.80 and 0.95 quantiles remained positive but were not statistically significant. Conclusion: QR revealed a broadly positive FAFH‐BMI association across the BMI distribution, with statistically significant associations through the 70th percentile and less precise estimates in the upper tail. Although further research is needed to better understand these associations, our findings highlight the importance of considering distributional patterns and statistical uncertainty when evaluating the impact of FAFH consumption on BMI and designing public health interventions.
- Open Access
- 24 July 2025
- Published by Springer Nature in Discover Data
- Vol. 3 (1) , 1-16
A new generalized inverse Gaussian distribution with Bayesian estimators
- 24 July 2025
- Published by Springer Nature in Discover Data
- Vol. 3 (1) , 1-16
Open AccessA new generalized inverse Gaussian distribution with Bayesian estimators
- Open Access
- 1 January 2025
- Published by School of Statistics, Renmin University of China in Journal of Data Science
- p. 1-20
Comparative Analysis of VADER and TextBlob on Financial News Headlines
Financial news headlines serve as a rich source of information on financial activities, offering a wealth of text that can provide insights into human behavior. One key analysis that can be conducted on this text is sentiment analysis. Despite extensive research over the years, sentiment analysis still faces challenges, particularly in handling internet slang, abbreviations, and emoticons commonly found on many websites that cover financial news headlines, including Bloomberg, Yahoo Finance, and Financial Times. This paper compares the performance of two sentiment analyzers—VADER and TextBlob—on financial news headlines from two countries: the USA (a well-developed economic nation) and Nepal (an underdeveloped economic nation). The collected headlines were manually classified into three categories (positive, negative, and neutral) from a financial perspective. The headlines were then cleaned and processed through the sentiment analyzers to compare their performance. The models’ performance is evaluated based on accuracy, sensitivity, specificity, and neutral specificity. Experimental results reveal that VADER performs better than TextBlob on both datasets. Additionally, both models perform better on financial news headlines from the USA than Nepal. These findings are further validated through statistical tests.
- 1 January 2025
- Published by School of Statistics, Renmin University of China in Journal of Data Science
- p. 1-20
Open AccessComparative Analysis of VADER and TextBlob on Financial News Headlines
Financial news headlines serve as a rich source of information on financial activities, offering a wealth of text that can provide insights into human behavior. One key analysis that can be conducted on this text is sentiment analysis. Despite extensive research over the years, sentiment analysis still faces challenges, particularly in handling internet slang, abbreviations, and emoticons commonly found on many websites that cover financial news headlines, including Bloomberg, Yahoo Finance, and Financial Times. This paper compares the performance of two sentiment analyzers—VADER and TextBlob—on financial news headlines from two countries: the USA (a well-developed economic nation) and Nepal (an underdeveloped economic nation). The collected headlines were manually classified into three categories (positive, negative, and neutral) from a financial perspective. The headlines were then cleaned and processed through the sentiment analyzers to compare their performance. The models’ performance is evaluated based on accuracy, sensitivity, specificity, and neutral specificity. Experimental results reveal that VADER performs better than TextBlob on both datasets. Additionally, both models perform better on financial news headlines from the USA than Nepal. These findings are further validated through statistical tests.
- 24 November 2024
- Published by Elsevier in Statistics & Probability Letters
- Vol. 218
Improving the Bickel–Rosenblatt global measure of deviation and goodness-of-fit test
- 24 November 2024
- Published by Elsevier in Statistics & Probability Letters
- Vol. 218
Improving the Bickel–Rosenblatt global measure of deviation and goodness-of-fit test
- Open Access
- 11 June 2024
- Published by MDPI AG in Econometrics
- Vol. 12 (2) , 16
Predicting the Direction of NEPSE Index Movement with News Headlines Using Machine Learning
Predicting stock market movement direction is a challenging task due to its fuzzy, chaotic, volatile, nonlinear, and complex nature. However, with advancements in artificial intelligence, abundant data availability, and improved computational capabilities, creating robust models capable of accurately predicting stock market movement is now feasible. This study aims to construct a predictive model using news headlines to predict stock market movement direction. It conducts a comparative analysis of five supervised classification machine learning algorithms—logistic regression (LR), support vector machine (SVM), random forest (RF), extreme gradient boosting (XGBoost), and artificial neural network (ANN)—to predict the next day’s movement direction of the close price of the Nepal Stock Exchange (NEPSE) index. Sentiment scores from news headlines are computed using the Valence Aware Dictionary for Sentiment Reasoning (VADER) and TextBlob sentiment analyzer. The models’ performance is evaluated based on sensitivity, specificity, accuracy, and the area under the receiver operating characteristic (ROC) curve (AUC). Experimental results reveal that all five models perform equally well when using sentiment scores from the TextBlob analyzer. Similarly, all models exhibit almost identical performance when using sentiment scores from the VADER analyzer, except for minor variations in AUC in SVM vs. LR and SVM vs. ANN. Moreover, models perform relatively better when using sentiment scores from the TextBlob analyzer compared to the VADER analyzer. These findings are further validated through statistical tests.
- 11 June 2024
- Published by MDPI AG in Econometrics
- Vol. 12 (2) , 16
Open AccessPredicting the Direction of NEPSE Index Movement with News Headlines Using Machine Learning
Predicting stock market movement direction is a challenging task due to its fuzzy, chaotic, volatile, nonlinear, and complex nature. However, with advancements in artificial intelligence, abundant data availability, and improved computational capabilities, creating robust models capable of accurately predicting stock market movement is now feasible. This study aims to construct a predictive model using news headlines to predict stock market movement direction. It conducts a comparative analysis of five supervised classification machine learning algorithms—logistic regression (LR), support vector machine (SVM), random forest (RF), extreme gradient boosting (XGBoost), and artificial neural network (ANN)—to predict the next day’s movement direction of the close price of the Nepal Stock Exchange (NEPSE) index. Sentiment scores from news headlines are computed using the Valence Aware Dictionary for Sentiment Reasoning (VADER) and TextBlob sentiment analyzer. The models’ performance is evaluated based on sensitivity, specificity, accuracy, and the area under the receiver operating characteristic (ROC) curve (AUC). Experimental results reveal that all five models perform equally well when using sentiment scores from the TextBlob analyzer. Similarly, all models exhibit almost identical performance when using sentiment scores from the VADER analyzer, except for minor variations in AUC in SVM vs. LR and SVM vs. ANN. Moreover, models perform relatively better when using sentiment scores from the TextBlob analyzer compared to the VADER analyzer. These findings are further validated through statistical tests.
- 4 June 2024
- Published by Springer Nature in Annals of Data Science
- Vol. 12 (4) , 1113-1156
Real Estate Market Prediction Using Deep Learning Models
- 4 June 2024
- Published by Springer Nature in Annals of Data Science
- Vol. 12 (4) , 1113-1156
Real Estate Market Prediction Using Deep Learning Models
- Open Access
Deep-SDM: A Unified Computational Framework for Sequential Data Modeling Using Deep Learning Models
Deep-SDM is a unified layer framework built on TensorFlow/Keras and written in Python 3.12. The framework aligns with the modular engineering principles for the design and development strategy. Transparency, reproducibility, and recombinability are the framework’s primary design criteria. The platform can extract valuable insights from numerical and text data and utilize them to predict future values by implementing long short-term memory (LSTM), gated recurrent unit (GRU), and convolution neural network (CNN). Its end-to-end machine learning pipeline involves a sequence of tasks, including data exploration, input preparation, model construction, hyperparameter tuning, performance evaluations, visualization of results, and statistical analysis. The complete process is systematic and carefully organized, from data import to model selection, encapsulating it into a unified whole. The multiple subroutines work together to provide a user-friendly and conducive pipeline that is easy to use. We utilized the Deep-SDM framework to predict the Nepal Stock Exchange (NEPSE) index to validate its reproducibility and robustness and observed impressive results.
Open AccessDeep-SDM: A Unified Computational Framework for Sequential Data Modeling Using Deep Learning Models
Deep-SDM is a unified layer framework built on TensorFlow/Keras and written in Python 3.12. The framework aligns with the modular engineering principles for the design and development strategy. Transparency, reproducibility, and recombinability are the framework’s primary design criteria. The platform can extract valuable insights from numerical and text data and utilize them to predict future values by implementing long short-term memory (LSTM), gated recurrent unit (GRU), and convolution neural network (CNN). Its end-to-end machine learning pipeline involves a sequence of tasks, including data exploration, input preparation, model construction, hyperparameter tuning, performance evaluations, visualization of results, and statistical analysis. The complete process is systematic and carefully organized, from data import to model selection, encapsulating it into a unified whole. The multiple subroutines work together to provide a user-friendly and conducive pipeline that is easy to use. We utilized the Deep-SDM framework to predict the Nepal Stock Exchange (NEPSE) index to validate its reproducibility and robustness and observed impressive results.
- Open Access
- 25 July 2022
- Published by Elsevier in Machine Learning with Applications
- Vol. 9
Predicting NEPSE index price using deep learning models
- 25 July 2022
- Published by Elsevier in Machine Learning with Applications
- Vol. 9
Open AccessPredicting NEPSE index price using deep learning models
- Open Access
- 1 January 2022
- Published by Elsevier in SSRN Electronic Journal
Predicting Nepse Index Price Using Deep Learning Models
Stock price prediction is a prevalent research field in both industry and academia. There is a pressing demand to develop a prediction model that captures the pattern of the financial activities with high precision to make an informed decision. Stock price prediction is challenging due to the complex, incomplete, fuzzy, nonlinear, and volatile nature of financial data. However, developing a robust model is possible due to advancements in artificial intelligence, availability of large-scale data, and increased access to computational capability.
- 1 January 2022
- Published by Elsevier in SSRN Electronic Journal
Open AccessPredicting Nepse Index Price Using Deep Learning Models
Stock price prediction is a prevalent research field in both industry and academia. There is a pressing demand to develop a prediction model that captures the pattern of the financial activities with high precision to make an informed decision. Stock price prediction is challenging due to the complex, incomplete, fuzzy, nonlinear, and volatile nature of financial data. However, developing a robust model is possible due to advancements in artificial intelligence, availability of large-scale data, and increased access to computational capability.