Search results
1 – 10 of 120Harish Kundra, Sudhir Sharma, P. Nancy and Dasari Kalyani
Bitcoin has indeed been universally acknowledged as an investment asset in recent decades, after the boom-and-bust of cryptocurrency values. Because of its extreme volatility, it…
Abstract
Purpose
Bitcoin has indeed been universally acknowledged as an investment asset in recent decades, after the boom-and-bust of cryptocurrency values. Because of its extreme volatility, it requires accurate forecasts to build economic decisions. Although prior research has utilized machine learning to improve Bitcoin price prediction accuracy, few have looked into the plausibility of using multiple modeling approaches on datasets containing varying data types and volumetric attributes. Thus, this paper aims to propose a bitcoin price prediction model.
Design/methodology/approach
In this research work, a bitcoin price prediction model is introduced by following three major phases: Data collection, feature extraction and price prediction. Initially, the collected Bitcoin time-series data will be preprocessed and the original features will be extracted. To make this work good-fit with a high level of accuracy, we have been extracting the second order technical indicator based features like average true range (ATR), modified-exponential moving average (M-EMA), relative strength index and rate of change and proposed decomposed inter-day difference. Subsequently, these extracted features along with the original features will be subjected to prediction phase, where the prediction of bitcoin price value is attained precisely from the constructed two-level ensemble classifier. The two-level ensemble classifier will be the amalgamation of two fabulous classifiers: optimized convolutional neural network (CNN) and bidirectional long/short-term memory (BiLSTM). To cope up with the volatility characteristics of bitcoin prices, it is planned to fine-tune the weight parameter of CNN by a new hybrid optimization model. The proposed hybrid optimization model referred as black widow updated rain optimization (BWURO) model will be conceptual blended of rain optimization algorithm and black widow optimization algorithm.
Findings
The proposed work is compared over the existing models in terms of convergence, MAE, MAPE, MARE, MSE, MSPE, MRSE, Root Mean Square Error (RMSE), RMSPE and RMSRE, respectively. These evaluations have been conducted for both algorithmic performance as well as classifier performance. At LP = 50, the MAE of the proposed work is 0.023372, which is 59.8%, 72.2%, 62.14% and 64.08% better than BWURO + Bi-LSTM, CNN + BWURO, NN + BWURO and SVM + BWURO, respectively.
Originality/value
In this research work, a new modified EMA feature is extracted, which makes the bitcoin price prediction more efficient. In this research work, a two-level ensemble classifier is constructed in the price prediction phase by blending the Bi-LSTM and optimized CNN, respectively. To deal with the volatility of bitcoin values, a novel hybrid optimization model is used to fine-tune the weight parameter of CNN.
Details
Keywords
Shrawan Kumar Trivedi, Amrinder Singh and Somesh Kumar Malhotra
There is a need to predict whether the consumers liked the stay in the hotel rooms or not, and to remove the aspects the customers did not like. Many customers leave a review…
Abstract
Purpose
There is a need to predict whether the consumers liked the stay in the hotel rooms or not, and to remove the aspects the customers did not like. Many customers leave a review after staying in the hotel. These reviews are mostly given on the website used to book the hotel. These reviews can be considered as a valuable data, which can be analyzed to provide better services in the hotels. The purpose of this study is to use machine learning techniques for analyzing the given data to determine different sentiment polarities of the consumers.
Design/methodology/approach
Reviews given by hotel customers on the Tripadvisor website, which were made available publicly by Kaggle. Out of 10,000 reviews in the data, a sample of 3,000 negative polarity reviews (customers with bad experiences) in the hotel and 3,000 positive polarity reviews (customers with good experiences) in the hotel is taken to prepare data set. The two-stage feature selection was applied, which first involved greedy selection method and then wrapper method to generate 37 most relevant features. An improved stacked decision tree (ISD) classifier) is built, which is further compared with state-of-the-art machine learning algorithms. All the tests are done using R-Studio.
Findings
The results showed that the new model was satisfactory overall with 80.77% accuracy after doing in-depth study with 50–50 split, 80.74% accuracy for 66–34 split and 80.25% accuracy for 80–20 split, when predicting the nature of the customers’ experience in the hotel, i.e. whether they are positive or negative.
Research limitations/implications
The implication of this research is to provide a showcase of how we can predict the polarity of potentially popular reviews. This helps the authors’ perspective to help the hotel industries to take corrective measures for the betterment of business and to promote useful positive reviews. This study also has some limitations like only English reviews are considered. This study was restricted to the data from trip-adviser website; however, a new data may be generated to test the credibility of the model. Only aspect-based sentiment classification is considered in this study.
Originality/value
Stacking machine learning techniques have been proposed. At first, state-of-the-art classifiers are tested on the given data, and then, three best performing classifiers (decision tree C5.0, random forest and support vector machine) are taken to build stack and to create ISD classifier.
Details
Keywords
Spam emails classification using data mining and machine learning approaches has enticed the researchers' attention duo to its obvious positive impact in protecting internet…
Abstract
Spam emails classification using data mining and machine learning approaches has enticed the researchers' attention duo to its obvious positive impact in protecting internet users. Several features can be used for creating data mining and machine learning based spam classification models. Yet, spammers know that the longer they will use the same set of features for tricking email users the more probably the anti-spam parties might develop tools for combating this kind of annoying email messages. Spammers, so, adapt by continuously reforming the group of features utilized for composing spam emails. For that reason, even though traditional classification methods possess sound classification results, they were ineffective for lifelong classification of spam emails duo to the fact that they might be prone to the so-called “Concept Drift”. In the current study, an enhanced model is proposed for ensuring lifelong spam classification model. For the evaluation purposes, the overall performance of the suggested model is contrasted against various other stream mining classification techniques. The results proved the success of the suggested model as a lifelong spam emails classification method.
Details
Keywords
Koraljka Golub, Osma Suominen, Ahmed Taiye Mohammed, Harriet Aagaard and Olof Osterman
In order to estimate the value of semi-automated subject indexing in operative library catalogues, the study aimed to investigate five different automated implementations of an…
Abstract
Purpose
In order to estimate the value of semi-automated subject indexing in operative library catalogues, the study aimed to investigate five different automated implementations of an open source software package on a large set of Swedish union catalogue metadata records, with Dewey Decimal Classification (DDC) as the target classification system. It also aimed to contribute to the body of research on aboutness and related challenges in automated subject indexing and evaluation.
Design/methodology/approach
On a sample of over 230,000 records with close to 12,000 distinct DDC classes, an open source tool Annif, developed by the National Library of Finland, was applied in the following implementations: lexical algorithm, support vector classifier, fastText, Omikuji Bonsai and an ensemble approach combing the former four. A qualitative study involving two senior catalogue librarians and three students of library and information studies was also conducted to investigate the value and inter-rater agreement of automatically assigned classes, on a sample of 60 records.
Findings
The best results were achieved using the ensemble approach that achieved 66.82% accuracy on the three-digit DDC classification task. The qualitative study confirmed earlier studies reporting low inter-rater agreement but also pointed to the potential value of automatically assigned classes as additional access points in information retrieval.
Originality/value
The paper presents an extensive study of automated classification in an operative library catalogue, accompanied by a qualitative study of automated classes. It demonstrates the value of applying semi-automated indexing in operative information retrieval systems.
Details
Keywords
Fatima-Zahrae Nakach, Hasnae Zerouaoui and Ali Idri
Histopathology biopsy imaging is currently the gold standard for the diagnosis of breast cancer in clinical practice. Pathologists examine the images at various magnifications to…
Abstract
Purpose
Histopathology biopsy imaging is currently the gold standard for the diagnosis of breast cancer in clinical practice. Pathologists examine the images at various magnifications to identify the type of tumor because if only one magnification is taken into account, the decision may not be accurate. This study explores the performance of transfer learning and late fusion to construct multi-scale ensembles that fuse different magnification-specific deep learning models for the binary classification of breast tumor slides.
Design/methodology/approach
Three pretrained deep learning techniques (DenseNet 201, MobileNet v2 and Inception v3) were used to classify breast tumor images over the four magnification factors of the Breast Cancer Histopathological Image Classification dataset (40×, 100×, 200× and 400×). To fuse the predictions of the models trained on different magnification factors, different aggregators were used, including weighted voting and seven meta-classifiers trained on slide predictions using class labels and the probabilities assigned to each class. The best cluster of the outperforming models was chosen using the Scott–Knott statistical test, and the top models were ranked using the Borda count voting system.
Findings
This study recommends the use of transfer learning and late fusion for histopathological breast cancer image classification by constructing multi-magnification ensembles because they perform better than models trained on each magnification separately.
Originality/value
The best multi-scale ensembles outperformed state-of-the-art integrated models and achieved an accuracy mean value of 98.82 per cent, precision of 98.46 per cent, recall of 100 per cent and F1-score of 99.20 per cent.
Details
Keywords
Huaxiang Song, Chai Wei and Zhou Yong
The paper aims to tackle the classification of Remote Sensing Images (RSIs), which presents a significant challenge for computer algorithms due to the inherent characteristics of…
Abstract
Purpose
The paper aims to tackle the classification of Remote Sensing Images (RSIs), which presents a significant challenge for computer algorithms due to the inherent characteristics of clustered ground objects and noisy backgrounds. Recent research typically leverages larger volume models to achieve advanced performance. However, the operating environments of remote sensing commonly cannot provide unconstrained computational and storage resources. It requires lightweight algorithms with exceptional generalization capabilities.
Design/methodology/approach
This study introduces an efficient knowledge distillation (KD) method to build a lightweight yet precise convolutional neural network (CNN) classifier. This method also aims to substantially decrease the training time expenses commonly linked with traditional KD techniques. This approach entails extensive alterations to both the model training framework and the distillation process, each tailored to the unique characteristics of RSIs. In particular, this study establishes a robust ensemble teacher by independently training two CNN models using a customized, efficient training algorithm. Following this, this study modifies a KD loss function to mitigate the suppression of non-target category predictions, which are essential for capturing the inter- and intra-similarity of RSIs.
Findings
This study validated the student model, termed KD-enhanced network (KDE-Net), obtained through the KD process on three benchmark RSI data sets. The KDE-Net surpasses 42 other state-of-the-art methods in the literature published from 2020 to 2023. Compared to the top-ranked method’s performance on the challenging NWPU45 data set, KDE-Net demonstrated a noticeable 0.4% increase in overall accuracy with a significant 88% reduction in parameters. Meanwhile, this study’s reformed KD framework significantly enhances the knowledge transfer speed by at least three times.
Originality/value
This study illustrates that the logit-based KD technique can effectively develop lightweight CNN classifiers for RSI classification without substantial sacrifices in computation and storage costs. Compared to neural architecture search or other methods aiming to provide lightweight solutions, this study’s KDE-Net, based on the inherent characteristics of RSIs, is currently more efficient in constructing accurate yet lightweight classifiers for RSI classification.
Details
Keywords
Thanh-Nghi Do and Minh-Thu Tran-Nguyen
This study aims to propose novel edge device-tailored federated learning algorithms of local classifiers (stochastic gradient descent, support vector machines), namely, FL-lSGD…
Abstract
Purpose
This study aims to propose novel edge device-tailored federated learning algorithms of local classifiers (stochastic gradient descent, support vector machines), namely, FL-lSGD and FL-lSVM. These algorithms are designed to address the challenge of large-scale ImageNet classification.
Design/methodology/approach
The authors’ FL-lSGD and FL-lSVM trains in a parallel and incremental manner to build an ensemble local classifier on Raspberry Pis without requiring data exchange. The algorithms load small data blocks of the local training subset stored on the Raspberry Pi sequentially to train the local classifiers. The data block is split into k partitions using the k-means algorithm, and models are trained in parallel on each data partition to enable local data classification.
Findings
Empirical test results on the ImageNet data set show that the authors’ FL-lSGD and FL-lSVM algorithms with 4 Raspberry Pis (Quad core Cortex-A72, ARM v8, 64-bit SoC @ 1.5GHz, 4GB RAM) are faster than the state-of-the-art LIBLINEAR algorithm run on a PC (Intel(R) Core i7-4790 CPU, 3.6 GHz, 4 cores, 32GB RAM).
Originality/value
Efficiently addressing the challenge of large-scale ImageNet classification, the authors’ novel federated learning algorithms of local classifiers have been tailored to work on the Raspberry Pi. These algorithms can handle 1,281,167 images and 1,000 classes effectively.
Details
Keywords
Chong Wu, Xiaofang Chen and Yongjie Jiang
While the Chinese securities market is booming, the phenomenon of listed companies falling into financial distress is also emerging, which affects the operation and development of…
Abstract
Purpose
While the Chinese securities market is booming, the phenomenon of listed companies falling into financial distress is also emerging, which affects the operation and development of enterprises and also jeopardizes the interests of investors. Therefore, it is important to understand how to accurately and reasonably predict the financial distress of enterprises.
Design/methodology/approach
In the present study, ensemble feature selection (EFS) and improved stacking were used for financial distress prediction (FDP). Mutual information, analysis of variance (ANOVA), random forest (RF), genetic algorithms, and recursive feature elimination (RFE) were chosen for EFS to select features. Since there may be missing information when feeding the results of the base learner directly into the meta-learner, the features with high importance were fed into the meta-learner together. A screening layer was added to select the meta-learner with better performance. Finally, Optima hyperparameters were used for parameter tuning by the learners.
Findings
An empirical study was conducted with a sample of A-share listed companies in China. The F1-score of the model constructed using the features screened by EFS reached 84.55%, representing an improvement of 4.37% compared to the original features. To verify the effectiveness of improved stacking, benchmark model comparison experiments were conducted. Compared to the original stacking model, the accuracy of the improved stacking model was improved by 0.44%, and the F1-score was improved by 0.51%. In addition, the improved stacking model had the highest area under the curve (AUC) value (0.905) among all the compared models.
Originality/value
Compared to previous models, the proposed FDP model has better performance, thus bridging the research gap of feature selection. The present study provides new ideas for stacking improvement research and a reference for subsequent research in this field.
Details
Keywords
Christian Nnaemeka Egwim, Hafiz Alaka, Youlu Pan, Habeeb Balogun, Saheed Ajayi, Abdul Hye and Oluwapelumi Oluwaseun Egunjobi
The study aims to develop a multilayer high-effective ensemble of ensembles predictive model (stacking ensemble) using several hyperparameter optimized ensemble machine learning…
Abstract
Purpose
The study aims to develop a multilayer high-effective ensemble of ensembles predictive model (stacking ensemble) using several hyperparameter optimized ensemble machine learning (ML) methods (bagging and boosting ensembles) trained with high-volume data points retrieved from Internet of Things (IoT) emission sensors, time-corresponding meteorology and traffic data.
Design/methodology/approach
For a start, the study experimented big data hypothesis theory by developing sample ensemble predictive models on different data sample sizes and compared their results. Second, it developed a standalone model and several bagging and boosting ensemble models and compared their results. Finally, it used the best performing bagging and boosting predictive models as input estimators to develop a novel multilayer high-effective stacking ensemble predictive model.
Findings
Results proved data size to be one of the main determinants to ensemble ML predictive power. Second, it proved that, as compared to using a single algorithm, the cumulative result from ensemble ML algorithms is usually always better in terms of predicted accuracy. Finally, it proved stacking ensemble to be a better model for predicting PM2.5 concentration level than bagging and boosting ensemble models.
Research limitations/implications
A limitation of this study is the trade-off between performance of this novel model and the computational time required to train it. Whether this gap can be closed remains an open research question. As a result, future research should attempt to close this gap. Also, future studies can integrate this novel model to a personal air quality messaging system to inform public of pollution levels and improve public access to air quality forecast.
Practical implications
The outcome of this study will aid the public to proactively identify highly polluted areas thus potentially reducing pollution-associated/ triggered COVID-19 (and other lung diseases) deaths/ complications/ transmission by encouraging avoidance behavior and support informed decision to lock down by government bodies when integrated into an air pollution monitoring system
Originality/value
This study fills a gap in literature by providing a justification for selecting appropriate ensemble ML algorithms for PM2.5 concentration level predictive modeling. Second, it contributes to the big data hypothesis theory, which suggests that data size is one of the most important factors of ML predictive capability. Third, it supports the premise that when using ensemble ML algorithms, the cumulative output is usually always better in terms of predicted accuracy than using a single algorithm. Finally developing a novel multilayer high-performant hyperparameter optimized ensemble of ensembles predictive model that can accurately predict PM2.5 concentration levels with improved model interpretability and enhanced generalizability, as well as the provision of a novel databank of historic pollution data from IoT emission sensors that can be purchased for research, consultancy and policymaking.
Details
Keywords
Jamil Jaber, Rami S. Alkhawaldeh and Ibrahim N. Khatatbeh
This study aims to develop a novel approach for predicting default risk in bancassurance, which plays a crucial role in the relationship between interest rates in banks and…
Abstract
Purpose
This study aims to develop a novel approach for predicting default risk in bancassurance, which plays a crucial role in the relationship between interest rates in banks and premium rates in insurance companies. The proposed method aims to improve default risk predictions and assist with client segmentation in the banking system.
Design/methodology/approach
This research introduces the group method of data handling (GMDH) technique and a diversified classifier ensemble based on GMDH (dce-GMDH) for predicting default risk. The data set comprises information from 30,000 credit card clients of a large bank in Taiwan, with the output variable being a dummy variable distinguishing between default risk (0) and non-default risk (1), whereas the input variables comprise 23 distinct features characterizing each customer.
Findings
The results of this study show promising outcomes, highlighting the usefulness of the proposed technique for bancassurance and client segmentation. Remarkably, the dce-GMDH model consistently outperforms the conventional GMDH model, demonstrating its superiority in predicting default risk based on various error criteria.
Originality/value
This study presents a unique approach to predicting default risk in bancassurance by using the GMDH and dce-GMDH neural network models. The proposed method offers a valuable contribution to the field by showcasing improved accuracy and enhanced applicability within the banking sector, offering valuable insights and potential avenues for further exploration.
Details