Browse Items (14421 total)
Sort by:
-
Predicting cryptocurrency prices model using a stacked sparse autoencoder and Bayesian optimization
In recent years, digital currencies, also known as cybercash, digital money, and electronic money, have gained significant attention from researchers and investors alike. Cryptocurrency has emerged as a result of advancements in financial technology and has presented a unique opening for research in the field. However, predicting the prices of cryptocurrencies is a challenging task due to their dynamic and volatile nature. This study aims to address this challenge by introducing a new prediction model called Bayesian optimization with stacked sparse autoencoder-based cryptocurrency price prediction (BOSSAE-CPP). The main objective of this model is to effectively predict the prices of cryptocurrencies. To achieve this goal, the BOSSAE-CPP model employs a stacked sparse autoencoder (SSAE) for the prediction process and resulting in improved predictive outcomes. The results were compared to other models, and it was found that the BOSSAE-CPP model performed significantly better. 2023, IGI Global. -
Predicting customer churn: A systematic literature review
Churn prediction is an active topic for research and machine learning approaches have made significant contributions in this domain. Models built to address customer churn, aim to identify customers who are at a high risk of terminating services offered by a company. Hence, an effective machine learning model indirectly contributes to the revenue growth of an organization, by identifying at risk customers, well in advance. This improves the success rate of retention campaigns and reduces costs associated with churn. The aim of this study is to explore the state-of-the-art machine learning techniques used in churn prediction. A systematic literature review, that is driven by 5 research questions and rigorous quality assessment criteria, is presented. There are 38 primary studies that are selected out of 420 studies published between 2018 and 2021. The review identifies popular machine learning techniques used in churn prediction and provides directions for future research. Firstly, the study finds that churn models lack generalization capability across industry domains. Hence, it identifies a need for researchers to explore techniques that extend beyond model experimentation, to improve efficiency of classifiers across domains. Secondly, it is observed that the traditional approaches to churn prediction depend significantly on demographic, product-usage, and revenue features alone. However, recent papers have integrated social network analysis-related features in churn models and achieved satisfactory results. Furthermore, there is a lack of scientific work that utilizes information-rich content of customer-company-interaction instances via email, chat conversations and other means. This area is the least explored. Thirdly, there is scope to investigate the effect of hybrid sampling strategies on model performance. This has not been extensively evaluated in literature. Lastly, there is no formal guideline on correct evaluation parameters to be used for models applied on imbalanced churn datasets. This is a grey area that requires greater attention. 2022 Taru Publications. -
Predicting electric vehicle performance metrics using a convolution neural network-gated recurrent unit-attention based deep learning architecture
The indicators of electric vehicle performance such as state of charge (SOC), remaining useful life (RUL), and charge demand need to be accurately forecasted to ensure maximum energy control and battery life. The models used are usually not able to capture the spatial and temporal correlation of battery data and be robust to the presence of noisy measurements. In this study, we model a sequential attention-based deep learning structure with convolutional neural networks, gated recurrent units, and an attention mechanism that can ultimately understand the local features, temporal relationships, and dynamic significance of various features in sequential battery data. The hybrid architecture of this model allows it to extract local spatial features, long-term sequential dependencies and dynamically find the importance of the critical time steps. We also develop a hybrid loss that is an accumulation of Huber loss and Mean Squared Error, which is much more resilient to outliers and at the same time has high prediction accuracy. It is experimentally proven that the proposed model has R2 values of 0.9575, 0.9558, and 0.9199 on SOC, RUL, and charge demand, respectively, which are better than the current single-architecture methods. 2026 The Authors -
Predicting emotional intelligence, creative performance and knowledge management in higher education using multiple regression
Higher education institutions are paramount in emerging nations like India. Post-globalisation, India witnessed the growth of HEIs, especially in the private sector. However, today most of the institutions are struggling for their existence. One of the most vital reasons for such a staggering performance is the absence of creativity. It will not be an exaggeration to say that the present era is the era of creativity and performance and organisations that cant perform are bound to perish. Creativity can be nurtured and yield success only if it is supported by the emotional intelligence (EI) of the employees and knowledge management (KM) processes. The current paper explored the nexus between emotional intelligence, knowledge management processes and creative performance in HEIs in India and implied that though emotional intelligence affects creative performance, the impact gets manifolded in the presence of the knowledge management process. Copyright 2025 Inderscience Enterprises Ltd. -
Predicting Employee Attrition Using Machine Learning Algorithms
Employees are considered the foundation of any organization. Due to their importance, the Human resources department implements various policies to sustain them. Yet the attrition rate in any organization is increasing yearly. The attrition rate signifies the number of employees who leaves a firm without being replaced. It is regarded as a well-known issue that requires the administration to make the best choices to retain highly competent staff. It is interesting to note that artificial intelligence is frequently used as a successful technique for foreseeing such an issue. This review paper aims to study the different machine learning approaches that predict employee attrition and factors influencing an employee to attrite from an organization. A Hybrid model comprising the various ensemble models is proposed to predict attrition at its earliest. The forecasted attrition model aids in not only taking preventive action but also in improving recruiting choices and rewarding top performers who contribute to the company's success. 2022 IEEE. -
Predicting energy source diversification in emerging Asia: The role of global supply chain pressure
This study investigates energy diversification trends in six Emerging Asian countries from 1998 to 2021 while exploring the predicting effects of the global supply chain pressure, total investment, innovation, economic growth, and globalisation on energy diversification. This study considers the Kernel-Based Regularized Least Squares (KRLS) estimations and prediction models (Adam and Stochastic Gradient Descent optimisers). The impacts of global supply chain pressure and total investment on energy diversification are positive. Innovation also emerges as crucial factor to enhance energy diversification. Deeper integration into the global economy (globalisation) and economic growth strengthen energy diversification. The study underscores the importance of tailored policies, advocating for investments in innovation, targeted total investment, and inclusive growth strategies to address energy diversification in emerging Asian countries. 2024 Elsevier B.V. -
Predicting financial asset prices with neural network: a comparative study of neural networks effectiveness in financial decision-making
Investing requires deep knowledge of complex financial markets, making it incredibly tedious to predict inflation and deflation. Predictive conventional models like ARIMA and GARCH do not accurately capture the non-linearity and volatility presented in financial datasets. This research examines the different forms of predictive assets, real estate, stocks, commodities, bonds, and cryptocurrency using Long Short-Term Memory (LSTM) Neural networks. The primary focus of this research is to assess the valuable prediction capabilities of LSTM across assets and its integration with financial decision-making. According to the empirical results, deep learning LSTM models give better outcomes with equities and gold, with the R2 indicator reaching over 99% alongside a low RMSE. LSTMs had an over 100% MPE prediction error rate for other assets during the test phase, making it harder to predict intensely volatile assets. The model's verification transfers residual autocorrelation, showing that it can enhance forecasting performance with detailed macroeconomic indicators and sentiment analysis data. Studies show that LSTMs are effective in high-frequency markets with non-linear price changes, but they require special attention to balance interpretability and overfitting. Despite the progress that has been achieved in utilizing neural networks for financial forecasting, hybrid models integrated with XAI are recommended to improve efficiency and real-world applicability. These results contribute to the growing domain of AI-powered finance by offering additional means for many investors, analysts, and decision-makers who wish to utilize data for market speculation. Dr. Aishwarya Nagarathinam et al. -
Predicting Financial Distress in India: A Deep Learning Approach
The present study examines the efficacy of deep learning models in predicting financial distress in India. For this purpose, the study employs three distinct architectures: Long Short-Term Memory (LSTM), Recurrent Neural Network (RNN), and Conventional Neural Network (CNN) models. Utilizing data from companies that filed for bankruptcy under the Insolvency and Bankruptcy Code 2016 for the period of 20162023, the study adopts a balanced sample approach to categorize them into distressed and non-distressed groups. Nineteen financial variables are utilized to predict financial distress. Python is used as the programming language, and Jupyter Notebook facilitates algorithm development. The findings reveal that the LSTM model, when compared to RNN and CNN, achieved 91% accuracy using parameters such as 8 LSTM units with tanh activation and a dense layer with sigmoid activation function, a batch size of 10, 50 epochs, RMSprop optimizer, and binary cross-entropy loss were used. The study suggests that deep learning presents a novel approach that can enhance performance in financial distress prediction studies. This study is believed to be the first to utilize deep learning models for financial distress prediction in India based on single-year data, offering valuable insights for financial institutions and investors seeking more effective risk management strategies. The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025. -
Predicting Financial Distress in India: A Deep Learning Approach
The present study examines the efficacy of deep learning models in predicting financial distress in India. For this purpose, the study employs three distinct architectures: Long Short-Term Memory (LSTM), Recurrent Neural Network (RNN), and Conventional Neural Network (CNN) models. Utilizing data from companies that filed for bankruptcy under the Insolvency and Bankruptcy Code 2016 for the period of 20162023, the study adopts a balanced sample approach to categorize them into distressed and non-distressed groups. Nineteen financial variables are utilized to predict financial distress. Python is used as the programming language, and Jupyter Notebook facilitates algorithm development. The findings reveal that the LSTM model, when compared to RNN and CNN, achieved 91% accuracy using parameters such as 8 LSTM units with tanh activation and a dense layer with sigmoid activation function, a batch size of 10, 50 epochs, RMSprop optimizer, and binary cross-entropy loss were used. The study suggests that deep learning presents a novel approach that can enhance performance in financial distress prediction studies. This study is believed to be the first to utilize deep learning models for financial distress prediction in India based on single-year data, offering valuable insights for financial institutions and investors seeking more effective risk management strategies. The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025. -
Predicting Financial Market Volatility Using Regression and Machine Learning Techniques
In standard Simple Linear Regression (SLR), one of the major assumptions is that the error terms have constant variance (homoscedasticity). However, this assumption is frequently violated in many real-world datasets, resulting in inefficient estimates and reduced predictive accuracy. To overcome this shortcoming, we propose a hybrid modeling platform that combines SLR with statistical and machine learning methods. The approach starts with SLR to identify the main linear relationship. Whenever residual diagnostics report the presence of heteroskedasticity, an Autoregressive Conditional Heteroskedasticity (ARCH) model is used to estimate time-varying variance. Such estimated variances are utilized in a Weighted Generalized Least Squares (WGLS) model, which stabilizes the error structure. Finally, to capture any remaining nonlinear patterns, an Artificial Neural Network (ANN) is applied on the residuals of the WGLS model. By layering these techniques, the hybrid framework improves both stability and predictive power. Simulation studies and empirical tests on Apple Inc. stock data confirmed that the hybrid framework yields reduced MAE and RMSE values and greater explanatory strength than individual approaches. 2025 IEEE. -
Predicting Football Players Market Value via Machine Learning
Football, arguably the most popular sport in the world, has become much more than just a sport, it is a multibillion-dollar industry with its center in Europe. Every year millions of euros are spent in transfer window to buy and sell players and a common theme that has been seen is players not living up to the price the clubs paid for them. This research aims to predict football players market values using machine learning techniques. Departing from traditional methods that broadly categorize players into positions like Goalkeeper, Defender, Midfielder, and Forward, this study provides a more nuanced approach by classifying players into specific roles such as Center-back, Full-back, Defensive Midfielder, Attacking Midfielder, and Winger. By incorporating performance metrics tailored to each position and weighing the performance indicators based on the relevance to that specific position, the research aims to provide a robust method to predict players market value within a negotiation tolerance range. Using data from the past three seasons, including detailed player performance statistics and contractual details, models were developed to assist clubs in making data-driven transfer decisions. Machine learning algorithms, including Random Forest Regressor and Light GBM, were utilized, with RMSE and R2 Score as evaluation metrics. Both algorithms demonstrated robust performance, with some positional models predicting market values within an acceptable error range of 312million, enabling clubs to negotiate transfer fees with greater precision based on empirical evidence of player performance. The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025. -
Predicting Graduate Admissions using Ensemble Machine Learning Techniques: A Comparative Study of Classifiers and Regressors
The goal of this research is to apply machine learning techniques to forecast a student's probability of being accepted into a graduate program. Applicants' GRE and TOEFL grades, university rankings, letters of recommendation, statements of purpose, cumulative grade point averages, and prior research experience are all included in the dataset utilized for this analysis. The goal is to calculate an applicant's expected acceptance rate. This study uses a combination of Classifiers and regressors. Different prediction models are contrasted in this study: Random Forest Classifier (RFC), Decision Tree Classifier (DTC), K-Neighbors Classifier (KNC), Support Vector Classifier (SVC), Gradient Boosting Classifier (GBC), Logistic regression (LR), Support vector Regressor (SVR), Random Forest Regressor(RFR), Gradient Boosting Regressor(GBR) and Decision Tree Regressor(DTR). Using these characteristics, the models are trained and evaluated. Evaluation criteria such as accuracy, kappa value, AUC-ROC, and confusion matrix are used to find the models' effectiveness. In order to determine which model performed the best, the assessment results are compared with one another. Based on study findings, the Gradient Boosting Classifier outperforms the other models tested by a significant margin (96 per cent). This model's AUC-ROC of 0.97 indicates it does a decent job at separating the positive and negative categories. 2023 IEEE. -
Predicting heart ailment in patients with varying number of features using data mining techniques
Data mining can be defined as a process of extracting unknown, verifiable and possibly helpful data from information. Among the various ailments, heart ailment is one of the primary reason behind death of individuals around the globe, hence in order to curb this, a detailed analysis is done using Data Mining. Many a times we limit ourselves with minimal attributes that are required to predict a patient with heart disease. By doing so we are missing on a lot of important attributes that are main causes for heart diseases. Hence, this research aims at considering almost all the important features affecting heart disease and performs the analysis step by step with minimal to maximum set of attributes using Data Mining techniques to predict heart ailments. The various classification methods used are Nae Bayes classifier, Random Forest and Random Tree which are applied on three datasets with different number of attributes but with a common class label. From the analysis performed, it shows that there is a gradual increase in prediction accuracies with the increase in the attributes irrespective of the classifiers used and Nae Bayes and Random Forest algorithms comparatively outperforms with these sets of data. 2019 Institute of Advanced Engineering and Science. -
Predicting Intention to Buy Organic Food during the COVID-19 Pandemic: A multi-group analysis based on the Health Belief Model
The ongoing COVID-19 pandemic has deeply affected physical and psychological health of people. It also had a huge impact on their dietary choices. This study specifically attempts to determine the impact of the constructs of health belief model on consumer purchase intention of organic food in the pandemic scenario. A survey was conducted among 413 Indian organic food consumers. The proposed hypotheses are tested by employing structural equation modeling. The findings highlight those perceived benefits is an important predictor of consumers behavioral intention to buy organic food, followed by cues to action and perceived threats. It is also found that consumers age moderates the impact of perceived threat and perceived barrier on consumers purchase intention, with a 22% difference in model prediction. In conclusion, the health belief model is found to be one of the most suitable models to predict consumer intention toward organic food purchase during the COVID-19 pandemic. 2022 Taylor & Francis Group, LLC. -
Predicting Job Risk from Artificial Intelligence in London Using Supervised Machine Learning Models
This study investigates the risk of job automation in London due to artificial intelligence (AI), applying supervised machine learning techniques to identify occupations most at risk. Leveraging a dataset encompassing job-specific features such as primary tasks, industry domains, and associated AI models, the research develops two predictive models. A Random Forest Classifier is used to categorize jobs as low, medium, or high automation risk, while a Linear Regression model estimates the proportion of each occupation's workload likely to be automated. The Random Forest model achieved a high accuracy rate of 97% in classifying job risk, indicating strong predictive capability. Meanwhile, the regression model explained 85% of the variance in the AI workload ratio, highlighting a significant relationship between job attributes and automation potential. These results suggest that job characteristics are reliable indicators of AI impact, particularly in routine, repetitive, and low-skilled roles that are more easily codified and replicated by algorithms. The findings align with broader economic theories such as creative destruction and technological waves, suggesting that AI not only displaces certain roles but also drives structural transformation within the labor market. By focusing on London, this study provides a localized understanding of how AI is reshaping employment patterns. It underscores the growing urgency for strategic workforce re-skilling and adaptive policy frameworks to mitigate negative outcomes and maximize opportunities presented by AI. Ultimately, this research contributes valuable insights into the interaction between AI technologies and employment, helping policymakers, employers, and educators anticipate change and prepare for a more resilient, inclusive labor market. 2025 IEEE. -
Predicting Liver Injury Risk from Chemical Properties and Drug Label Information Using Machine Learning Models
This research aims to create a drug-induced liver injury (DILI) severity prediction system based on machine learning to aid healthcare professionals in safety assessment. FDA's Liver Toxicity Knowledge Base supplied a drug dataset of 1042 drugs, and later, after pre-processing and API data extraction, each drug was defined by 16 chemical features such as molecular descriptors and pharmacokinetic properties. To improve uniformity and get quality input for training, data preparation involved correcting missing values, encoding categorical values, and normalising numerical data. Various machine learning models were trained and evaluated to forecast the levels of DILI severity, i.e., Random Forest, Gradient Boosting, and XGBoost. The importance of features was approximated for identifying the predictors that impacted the most. The best overall performance was recorded for XGBoost, and it had 81% accuracy when it was evaluated. Its acceptable discrimination was established for mild, moderate, and severe cases. The aptness of being applied to the medical sector is demonstrated by drastically lowering the principal misclassifications, especially from mild to severe. The application of machine learning in improving medicine safety assessment and reducing risks associated with pharmaceutical development is illustrated here. 2025 IEEE. -
Predicting Nitrogen Flavanol Index (NFI) in Mentha arvensis Using UAV Imaging and Machine Learning Techniques for Sustainable Agriculture
Crop growth monitoring at various growth stages is essential for optimizing agricultural inputs and enhancing crop yield. Nitrogen plays a critical role in plant development; however, its improper application can reduce productivity and, in the long term, degrade soil health. The aim of this study was to develop a non-invasive approach for nitrogen estimation through proxies (Nitrogen Flavanol Index) in Mentha arvensis using UAV-derived multispectral vegetation indices and machine learning models. Support Vector Regression, Random Forest, and Gradient Boosting were used to predict the Nitrogen Flavanol Index (NFI) across different growth stages. Among the tested models, Random Forest achieved the highest predictive accuracy (R2 = 0.86, RMSE = 0.32) at 75 days after planting (DAP), followed by Gradient Boosting (R2 = 0.75, RMSE = 0.43). Model performance was lowest during early growth stages (1530 DAP) but improved markedly from mid to late growth stages (4590 DAP). The findings highlight the significance of UAV-acquired data coupled with machine learning approaches for non-destructive nitrogen flavanol estimation, which can immensely contribute to improving real-time crop growth monitoring. 2025 by the authors. -
Predicting nitrous oxide contaminants in Cauvery basin using region-based convolutional neural network
Nitrous oxide (N2O) in riverbeds affects hydrological processes by contributing to the greenhouse effect, indicating poor water quality, disrupting biogeochemical cycling, and linking to eutrophication. Elevated N2O levels signal environmental issues, impacting aquatic life and necessitating precise forecasting for effective environmental management and reduced greenhouse gas emissions. Precisely forecasting nitrous oxide (N2O) emissions from riverbeds is paramount for effective environmental management, given its significant potency as a greenhouse gas. This study focuses on the difficulties related to spatial feature extraction and modeling accuracy in predicting N2O in riverbeds in Tamil Nadu. To address the obstacles, the research suggests utilizing the Deep Learning Based Prediction of Nitrous Oxide Contaminants (DL-PNOC), which studies the N2O contaminants in water using Region-based Convolutional Neural Network (RCNN) for spatial feature extraction, to predict nitrous oxide contaminants. The study is centered on the Cauvery River Basin located in Tamil Nadu, where the emission of N2O is a matter of environment. The outcomes encompass the specialized N2O contaminant model for riverbeds and the implementation of RCNN achieves precise N2O forecasting. The DL-PNOC approach combines a contaminant model with RCNN deep learning techniques to capture spatial characteristics and predict N2O pollutants accurately. Furthermore, using the River Bed Dynamics Simulator reinforces the dependability of the findings. The DL-PNOC approach has exhibited encouraging results, as evidenced by the following metrics: a high IoU of 88.66%, precision of 88.96%, recall of 90.03%, F1 score of 89.22%, and low RMSE and MAE values of 9.14% and 7.59%, respectively. The findings highlight the efficacy of the DL-PNOC approach in precisely forecasting N2O pollutants in river sediments. 2024 Elsevier B.V. -
Predicting of Credit Risk Using Machine Learning Algorithms
Credit risk management is one of the key processes for banks and is crucial to ensuring the banks stability and success. However, due to the need for more rigid forecasting models with strong mapping abilities, credit risk prediction has become challenging for the banking industry. Therefore, this paper attempts to predict commercial banks credit risk (CR) by using various machine learning algorithms. Machine learning algorithms, namely linear regression, KNN, SVR, DT, RF, XGB, and MLP, are compared with and without feature selection and feature extraction techniques to examine their prediction capabilities. Various determinants of credit risk (features) have been extracted to predict credit risk, and these features have been used to train machine learning models. Findings revealed that the decision tree algorithm had the highest performance, with the lowest mean absolute error (MSE) value of 0.1637 and the lowest root mean squared error (RMSE) value of 0.2158. The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024. -
Predicting of Open Source Software Component Reusability Level Using Object-Oriented Metrics by Taguchi Approach
Component-based software development (CBSD) is an efficient approach used by software developers to develop new software. The commercial off the shelf (COTS) and open-source software (OSS) are two styles to implement CBSD. The COTS provides the interface and depicts the black-box behavior, but does not support several software quality characteristics. On the other hard, OSS is a more efficient approach compared to COTS due to its source code availability. This research aims to identify the reusability level of OSS components from an online repository of OSS. The OSS components are classified based on Chidamber and Kemerer reusability metrics (CK-metrics). This paper proposed a mathematical model to establish the relationship between the reusability of CK-metrics. Reusability level of OSS component has been measured and most effective CK-metrics obtained by applying the Taguchi design and analysis of variance (ANOVA). The input parameters for the experimental design are evaluated based on the OSS repository. Performance analysis has been carried out based upon the interaction effect between the reusability of CK-metrics. Main effect plots are created to identify the most reusable component of the OSS. The genetic algorithm (GA) is used to predict the optimized value of the different control parameters. The results indicate that the OSS component reusability level is 0.698194. The reusability of software has a significant effect on the quality of software. The quality of software can be improved by increasing the reusability of software components. 2021 World Scientific Publishing Company.