Explore the vast range of titles available.

Purpose - In light of Bangladesh's economy, the goal of this study is to examine the \"Twin Deficit Hypothesis (TDH),\" which refers to a link between the budget deficit and the current account deficit. This study used yearly time series data from 1980 to 2020 to investigate the phenomena. Design/methodology/approach - A multivariate autoregressive distributive lag (ARDL) model has been presented for empirical investigation, with the ARDL bound test investigating the co-integration between the inadequacies. As some of the variables in the bound test lack co-integration, the study adds a multivariate vector autoregressive (VAR) model later on. Findings - With evidence of the result, the study supports the validation of twin deficit hypothesis in Bangladesh economy since both current account deficit and fiscal deficit affects each other significantly whereas Granger causality test confirms that fiscal deficit causes current account deficit but not the other way around. Practical implications - The government should maintain a restrictive monetary policy in order to stabilize the current account deficit. Originality/value - The novelty of this study is the incorporation of inflation, real exchange rate and GDP per capital to TDH that together form the basis for a macroeconomic snapshot of the economy.

In the domain of music information retrieval, emotion based classification is an active area of research. Emotion being a perceptual and subjective concept, the task is quite challenging. It is very difficult to design signal based descriptors to represent emotions. In this work deep leaning network is proposed and experiment is done with benchmark datasets namely, Soundtracks, Bi-Modal and MER_taffc. Experiment has also been done with hand crafted descriptor consisting of different time domain and spectral features, linear predictive coding and MFCC based features. Different classifiers like, neural network, support vector machine and random forest are tried. Although the combined feature set with neural network provides an optimal result for the datasets, but in general the performance of such approaches is limited. It is difficult to obtain a consistent feature set that works across the classifier and datasets. To get rid of the issue of feature design, deep learning based approach is followed. A convolutional neural network built around VGGNet and a novel post-processing technique are proposed. Proposed methodology provides substantial improvement of performance for the datasets. Comparison with other reported works on three different datasets also establishes the superiority of the proposed methodology. The improvement in performance has been substantiated by Z test.

This study is conducted to find the determinants of female labor force participation in Tangail, Bangladesh. The study examines the relationship of female labor force participation in this district with different factors like family type, education, training facilities and credit facilities. We purposively collect data from 300 females in Tangail district using the schedule method and use logistic regression. Later Hosmer and Lemeshow test, cooks distance test, and VIF test are used to diagnose whether the results are valid or not. Analysis shows that: (1) females from nuclear families are thirteen times more likely than those from joint families to be engaged in economic activities; (2) agriculture and service-oriented household’s women are less interested in joining the labor force than those from other income sources; (3) participation of women aged 55 to above is less in the labor force than those of others; (4) higher education, training facilities, and microcredit all have a positive impact on women's labor force participation because educated women participate much more than uneducated women; (5) females who receive training facilities are 32 times more likely , and women who receive microcredit are three times more likely to enter the labor force than their counterparts.

Background and Aims Surgical trauma initiates changes in central and peripheral nervous systems that need to be treated therapeutically to facilitate postoperative pain. The quality of postoperative analgesia is expected to affect clinical outcomes positively. Albeit optimal pain relief following cardiac surgery is often complex, researchers have tried to explore several techniques other than conventional ones during the last decade to find a unique analgesic method for postcardiac surgical patients. This study aims to find a unique analgesic approach that maximizes patient satisfaction after off‐pump coronary artery bypass graft (OPCABG) surgery. Methods The current study will compare the analgesic effect of upper thoracic epidural analgesia (TEA) with conventional analgesia after OPCAB graft surgery. For this, we will use a Quasi‐experimental study design. Patients admitted for coronary artery bypass graft (CABG) surgery will be assigned into two groups. The control group (conventional) will receive intravenous opioids and nonsteroidal anti‐inflammatory medications, and the study (case) group (TEA) will receive Inj. Bupivacaine 0.25% as an infusion through the epidural catheter. Physiologic parameters like hemodynamic and respiratory variables and pain scores will be recorded in predesigned format periodically. Results We expect to analyze a total of 130 consecutive off‐pump CABG surgery patients in Group A (Case, 65 patients) and Group B (Control, 65 patients). Study variables will be the visual analog scale score, hemodynamic parameters (heart rate, mean arterial pressure, and respiratory parameters (respiratory rate, PaO2, PaCO2, PEFR, FEV1). After data collection, the result will be analyzed and published in the public domain and in journals. Conclusion We expect thoracic epidural analgesia with local anesthetics will be a reliable postoperative analgesic option.

The study employs system GMM to derive the benchmark impact of foreign direct investment (FDI) on labor productivity (LP) as well as the interactive effects of FDI and productive capacity index (PCI) while dynamic panel threshold technique is used to determine the threshold level of PCI. Later to check the sensitivity test, pooled mean group (PMG) methodology is applied. Using panel data from 88 countries from 2000 to 2018, we examine the effect of FDI on LP contingent on the level of PCI across two economic sectors: tradables and nontradables. Applying novel PCI, the findings demonstrate that initially FDI exacerbates LP in the above two sectors, and the improvement in PCI from FDI diminishes this detrimental impact until a threshold of PCI, and then beyond that level, FDI enhances LP. Meanwhile, the latter benefit is larger in the tradable sector than in the nontradable sector, and this beneficial effect is amplified by increased FDI inflows. A set of robustness tests were performed to corroborate the findings. Notably, the internal mechanism of PCI’s eight indicators moderates the influence of FDI on LP. The study has a significant disadvantage as it spans 19 years (from 2000 to 2018). Since PCI data is not accessible before 2000 and after 2018, we have limited the investigation to this time period. The study applies novel PCI, a wide category index, to assess the host countries’ absorption capacity. Furthermore, according to our knowledge, this is the first attempt to assess the impact of FDI on LP in tradable and nontradable sectors, as well as to determine the PCI threshold level at which the effect of FDI switches direction. Policy implications of this study reveal that, while FDI may not directly increase LP, PCI-backed FDI growth may imply an increase in LP. The study presents policy considerations to enable the potential role of PCI indicators in facilitating the beneficial role of FDI on LP to increase competitiveness of the host economies.

In the transition to a society based on renewable energy, flexibility is important in balancing the energy supply as more intermittent sources like wind and solar are included in the energy mix. The storage-based hydropower systems are a renewable energy source that provides the needed flexibility since a hydropower plant can be started and stopped in minutes, and the reservoirs provide stored energy that can be utilized when the demand arises. Thereby, the hydropower plants can balance the variability in other energy sources, e.g., when there is no wind or when solar input is low. This need for increased flexibility has led research toward new hydropower turbines to provide larger ramping rates, more frequent starts and stops, and other system services. A possible drawback of the ramping operation of hydropower plants (often termed “hydropeaking”) are the adverse effects on the environment in receiving water bodies downstream of the power plant outlet, particularly when the hydropower outlets are in rivers. Rapid changes in flow can lead to stranding of fish and other biota during the shutdown of turbines and flushing of biota during the start of turbines. These effects can also be caused by other sudden episodes of water withdrawal, such as during accidental turbine shutdowns. The main objective of this study is to describe a method of designing the necessary volume of water required to mitigate a fast ramping turbine, and present the effect this has on the downstream river reach. We used a 2D hydraulic model to find the areas affected by hydropeaking operation and, furthermore, to define areas with a faster ramping rate than 13 cm/h which is used as a limit in Norwegian guidelines. Based on this, we developed a ramping regime that would prevent fast dewatering of critical areas and provide this as a basis for mitigating the effects of fast dewatering in the downstream river (River Nidelva in Norway was used as a test case). Furthermore, the effect of increasing the frequency of start–stop cycles was studied, and the proposed mitigation was evaluated for the new operational regime.

The Proposed work intends to automate the detection and classification of diabetic retinopathy from retinal fundus image which is very important in ophthalmology. Most of the existing methods use handcrafted features and those are fed to the classifier for detection and classification purpose. Recently convolutional neural network (CNN) is used for this classification problem but the architecture of CNN is manually designed. In this work, a genetic algorithm based technique is proposed to automatically determine the parameters of CNN and then the network is used for classification of diabetic retinopathy. The proposed CNN model consists of a series of convolution and pooling layer used for feature extraction. Finally support vector machine (SVM) is used for classification. Hyper-parameters like number of convolution and pooling layer, number of kernel and kernel size of convolution layer are determined by using the genetic algorithm. The proposed methodology is tested on publicly available Messidor dataset. The proposed method has achieved accuracy of 0.9867 and AUC of 0.9933. Experimental result shows that proposed auto-tuned CNN performs significantly better than the existing methods. Use of CNN takes away the burden of designing the image features and on the other hand genetic algorithm based methodology automates the design of CNN hyper-parameters.

Detection of mental states like stress/anxiety, mediation is a widely researched topic and is important for ensuring overall well-being of an individual. Several approaches have been reported in the literature for prediction or assessment of mental states. Recently, with advances in sensor technology, various physiological signals are being used by researchers for detecting mental states. In the present study, we have used a light weight deep convolutional neural network (CNN) for creating a mental state prediction model. The proposed detection model is created using publicly available WESAD dataset. The dataset contains electrocardiogram (ECG), galvanic skin response (GSR), skin temperature and electromyogram (EMG) signals recorded using a wearable device. Results show that for binary classification of stress vs no-stress condition our results are comparable with that reported in state-of-the-art machine learning/deep learning-based approaches. However, for three class classification of baseline vs stress vs amusement states, our model gives an accuracy of 90% which is much higher compared to that reported in the literature. In addition, we have also tried to classify various binary states like stress vs baseline , stress vs amusement and stress vs meditation conditions. The f1 score obtained for these classes are 0.96, 0.87 and 0.91, respectively, which are much higher than that reported in state-of-the-art literature using same dataset. Proposed light weight CNN-based mental state classification model is computationally less complex compared to other deep networks used by the researchers. Thus, it can be used for monitoring mental state successfully in real-life scenarios.

It has been shown in recent years that a range of optical diseases have early manifestation in retinal fundus images. It is becoming increasingly important to separate the regions of interest (RoI) upfront in the automated classification pipeline in order to ensure the alignment of the disease diagnosis with clinically relevant visual features. In this work, we introduce Pan-Ret, a semi-supervised framework which starts with locating the abnormalities in the biomedically relevant RoIs of a retinal image in an “annotation-agnostic” fashion. It does so by leveraging an anomaly detection setup using parallel autoencoders that are trained only on healthy population initially. Then, the anomalous images are separated based on the RoIs using a fully interpretable classifier like support vector machine (SVM). Experimental results show that the proposed approach yields an overall F1-score of 0.95 and 0.96 in detecting abnormalities on two different public datasets covering a diverse range of retinal diseases including diabetic retinopathy, hypertensive retinopathy, glaucoma, age-related macular degeneration, and several more in a staged manner. Thus, the work presents a milestone towards a pan-retinal disease diagnostic pipeline that can not only cater to the current set of disease classes, but has the capacity of adding further classes down the line. This is due to an anomaly detection style one-class learning setup of the deep autoencoder piece of the proposed pipeline, thus improving the generalizability of this approach compared to usual fully supervised competitors. This is also expected to increase the practical translational potential of Pan-Ret in a real-life scalable clinical setting. Graphical abstract Proposed Pan-Ret approach for detection of retinal abnormality using a set of clinically relevant RoIs from the retinal fundus images

The current study is focused on a novel methodology that removes the artifacts after automated segmentation of micrographs using some efficient image processing techniques from three dual-phase (DP) steel microstructures with varying martensite morphologies (islands, blocky and banded, and fine and fibrous martensite). Microstructures of DP steel consisting of martensite and ferrite phases were binarized by the Otsu thresholding algorithm, where the threshold value is automatically selected from the gray-level histogram. Subsequently, the artifact is removed by different modes, such as horizontally, vertically, and diagonal traverse manners, separately and in the combinations of all possible traverse manners, through AND operation. Optical micrographs were chosen for image analysis since the low-cost optical microscope is easily portable and effortlessly usable over a large area. The results have been validated with AxioVision commercial metallographic image analyzing software. The proposed scheme has successfully produced artifact-free binarized images of DP steel microstructures. Artifact-free microstructures provide the best quantification result, like the volume percent of the phases.