Purpose This paper aims to analyze and compare the energy forecasting models based on traditional statistical approaches and the currently prevalent Deep Learning (DL) approach. Design This experimental research utilized secondary data from the EIA website, consisting of monthly records of primary energy consumption in the U.S. residential sector from January 1973 to the present. Traditional methods for time series forecasting, such as ARIMA and Exponential Smoothing, were applied along with deep learning models like Multilayer Perceptron and Long Short-Term Memory networks. Forecast accuracy was measured using Root Mean Squared Error (RMSE) and Mean Absolute Percentage Error (MAPE). Findings LSTM, ARIMA, and Exponential Smoothing forecasts showed no statistically significant difference in their forecasts and outperformed the MLP model. Practical implications An Accurate forecasting model allows businesses to make informed decisions on market strategy, generation, transmission, and financial planning. It also reduces uncertainty and helps market participants prepare for possible risks, including supply chain interruptions or market swings. Originality/value This study compares conventional and deep learning models for long-term energy forecasting and provides insights into their performance and practical use.

The evolution of retail platforms and online marketplaces fundamentally transformed consumer expectations from their product purchases, particularly the last-mile delivery (LMD). In this era of urbanization, companies also strive for Sustainable Last Mile Delivery (S-LMD). Attaining the S-LMD has become challenging amid the environment concerns in online retailing. However, there are many obstacles in the way of attaining S-LMD, especially in a nation as diverse and dynamic as India. By thoroughly examining the challenges impeding the adoption of S-LMD in India's online retail industry, this study systematically identified, examined, and ranked the particular barriers of sustainable LMD. For ranking the S-LMD obstacles, we have applied Spherical Fuzzy Analytic Hierarchy Process (SF-AHP) technique. The SF-AHP has been a new tool that has been used for the very first time for identification and analysis of S-LMD related barriers. The extracted barriers were assessed by experts. Data was gathered with the help of expert interviews by academicians and business professionals throughout India. Five key categories of barriers were found by the study, which were then further categorised into 21 sub-barriers. After considering the global weights related to the full set of sub barriers it was found that Limited Sustainable Infrastructure, Traffic congestions and population density, High Initial Investment Cost for adopting S-LMD measures, Lack of Technical Knowledge on sustainable practices and infrastructure and Pressure for lower prices holds the top 5 positions among all the 21 sub-barriers which work as impediment for S-LMD. This study significantly contributes to the sustainable supply chain sector by bringing out the tactical recommendations and practical insights. Policymakers, logistics firms, and online corporations trying to build successful interventions for S-LMD in India and other emerging countries will find it to be a constructive resource.

The operational environment of software differs from the debugging environment. Therefore, the study explores the impact of irregular consumption by diverse users on the development cost of software, reliability of software, and release decisions. To accomplish this, a release model has been formulated considering logistic testing coverage-based reliability growth model built as a stochastic process including the error generation. The stochastic nature has been captured by considering the noise factor due to irregular fluctuations occurring during testing while usage uncertainty has been captured by introducing a constant parameter in the cost function of the operational phase. It is assumed that the testing phase cost is affected by the noise and the operation phase cost is affected by the severity of noise which is the result of uncertain usage by users. The model was evaluated against a real failure dataset. The release model creates a trade-off between software development cost, release timing, and reliability aspirations. This study contributes to software reliability literature and provides insights to practitioners to make software release decisions. The sensitivity analysis results give information about various aspects during the operational phase that affect the overall development cost.

The crucial part of the software development cycle is software maintenance. The demands included in the software management are fault fixes and request to change or bring a new feature. If priority is not given to these demands, then it may lead to customer dissatisfaction, inefficient planning, and software failure as well. Therefore, it is important to study the severity of the bug reports to maintain the efficiency of the software. Various research has been conducted in the past to predict the severity of the paper using text mining focusing only on the content of the bug reports. The sentiment of the user while reporting a bug also plays a vital role. In this study, we will be focusing on two aspects, that is, sentiment and content to improve the prediction. We propose a prediction model based on LDA to study the content aspect and emotion analysis to study the sentiment aspect. The model is validated on the datasets collected from the Eclipse project using Convolutional Neural Network (CNN). The results show that the CNN model effectively utilizes the content and sentiment aspect of the data to handle the severity prediction. CNN has weight sharing feature that decreases the number of parameters used for training. It also improves generalization and overfitting is avoided The Accuracy, Precision, Recall, and F-measure are improved when both aspects are taken into account rather than considering only content.

The efficiency and performance of a software application largely depend on the testing strategy adopted by the firm. Apart from the tools, techniques, and skills used for testing, the duration also plays an important influence in establishing software reliability. This defines the operational performance of the software. The testing duration decision is dependent on the failure behavior depicted by the software during the process of testing and the cost spent at various phases of development. In this paper, we study the multi-release model whose fault removal process is affected by random irregular fluctuations (white noise) and error generation phenomenon and determine the optimal time of testing for a multi-release-based software. The fault removal process is governed by the phenomenon of testing coverage which is affected by random fluctuations. The model shows encouraging results as it handles the stochastic property of the fault detection process. The optimal testing time is determined with the goal to minimize the expected development cost related to software while achieving the desired reliability levels of software for that release using a Genetic Algorithm. A real-life four-release fault dataset of tandem computers has been used to numerically demonstrate the methodology. It is observed through sensitivity analysis that the presence of white noise directly affects the cost and optimal testing duration. The potential to improve sensitivity, flexibility, early detection, discovering unsuspected patterns, and boost fault diagnosis is enhanced by collecting irregular fluctuations in the fault detection rate. This technology, by going beyond existing methodologies, has distinct advantages for detecting faults and can contribute to more dependable and efficient systems in a variety of domains.

Due to the increasing reliance on technology in nearly every industry over the past three decades, it has become necessary to evaluate the performance of a software product prior to its formal release in the market. The properties of a software application, such as its complexity and lines of code, are subject to change over time as a result of factors namely the testing environment, allocation of resources, testing efficiency, and testing team’s expertise. The assumption of constant Fault Detection Rate (FDR) may not accurately anticipate the potential number of bugs correctly. Keeping all these considerations in mind, a framework is developed to incorporate change point in the development of a testing effort-based Software Reliability Growth Model (SRGM) that takes into account the effect of application characteristics under both perfect and imperfect debugging settings. In addition, these outcomes are compared to the model without a change point. The proposed model is validated on two real-life software fault datasets. The results demonstrate that the proposed model performs better than the model without a change point.

The use of computers has been increasingly prevalent in our social lives in recent years. As a result, software engineers must create trustworthy software systems. Companies often release improved versions of the core program due to the constant demand and growing competition in the software field. A variety of growth models have been created to track and measure reliability by software managers and engineers. In the testing phase, the fault content and size of the software system increase, and consequently, the fault content found and eliminated during each debugging process decreases in comparison to the fault content present at the initial stage. In this scenario, we can consider the software fault detection process to be stochastic. The fault detection process is expressed in terms of testing coverage with random effects. In this study, we construct a testing coverage-based software model incorporating random effect with change point. We have used different testing coverage functions such as Exponential and Delayed S-shaped to study the effect of randomness. Further, multi-release planning for the proposed model has been studied and validated on the real-time failure dataset from Tandem Computers with four releases. Different performance measures and goodness-of-fit have been presented using a graphical representation.

The paper presents software reliability growth model (SRGM) with testing coverage which covers both detection as well as correction process under imperfect debugging. The estimation is done using feed forward back propagation artificial neural network (ANN). Many researchers have studied the importance of modelling fault detection instead of modelling fault correction. We have proposed generalised testing coverage model by adopting different testing coverage for both the processes. We have also compared proposed model with existing traditional models based on three failure datasets. Different performance criteria like goodness of fit, accuracy of the model, mean square error (MSE), and coefficient of determination (R-2) are evaluated for the datasets. The comparison results shows that the model proposed in this paper provides more efficient accuracy than the existing traditional models.

The most essential step during the development of the software is the testing procedure which makes the software dependable and efficient. During this procedure, the observation and rectification of the faults play a significant role in increasing the reliability of the software. Various Software Reliability growth models (SRGMs) with multiple assumptions were presented by various researchers to study the software’s reliability. It is well known that the fault observation/removal rate may get affected by irregular factors causing arbitrary effects. In this study, we aim to capture this irregular variation in fault observation/removal rate by expressing it in terms of testing coverage. The fault observation/removal process has been assumed as a stochastic process and modeled it using an Itô type of stochastic differential equation. Testing coverage enables software designers to check the software’s excellence and to see if any extra efforts are required to enhance reliability. In this paper, we have developed an SRGM based on testing coverage by introducing the concept of chang-epoint, error generation, and fault detection rate with irregular fluctuations. The error generation implies that during the testing procedure faults are not disclosed entirely and more faults get introduced. Later on, we focused on the idea of multi-release by considering four releases. We have estimated the parameters of the model by using the fault dataset for consecutive releases of Tandem Computers and validated the performance by evaluating the various goodness-of-fit criteria.

Organizations must build software that is extremely dependable due to the high expense of resolving errors, safety issues, and legal obligations. Software developers have created models for measuring and tracking the evolution of dependability in their products. Most of the proposed software reliability growth models takes fault detection into account throughout both the phases of testing and the operational as counting process. In addition, the size of the software system affects how many faults are discovered during testing and also how many are discovered and rectified during debugging relative to the fault content at the beginning of the testing period. So, in a situation like this, we may conceptualize the software fault detection process in terms of testing coverage as a stochastic process with a continuous state space. In this research, we offer an Ito-type stochastic differential equation-based ANN-based testing coverage software reliability growth model. The proposed approach has been tested and examined using real failure datasets from software projects. The suggested model that incorporates the idea of stochastic differential equations in testing coverage-based SRGM outperforms the current NHPP-based model.

In the past, various Software Reliability Growth Models (SRGMs) have been proposed using different parameters to improve software worthiness. Testing Coverage is one such parameter that has been studied in numerous models of software in the past and it has proved its influence on the reliability models. To sustain themselves in the market, software firms keep upgrading their software with new features or enhancements by rectifying previously reported faults. Also, there is an impact of the random effect on testing coverage during both the testing and operational phase. In this paper, we have proposed a Software reliability growth model based on testing coverage with random effect along with imperfect debugging. Later, the multi-release problem is presented for the proposed model. The proposed model is validated on the dataset from Tandem Computers. The results for each release of the models have been discussed based on the different performance criteria. The numerical results illustrate that models fit the failure data significantly.•The random effect in the testing coverage rate is handled using Stochastic Differential Equations (SDE).•Three testing coverage functions used are Exponential, Weibull, and S-shaped.•Four Releases of the software model has been presented.

Purpose The software development team reviews the testing phase to assess if the reliability growth of software is as per plan and requirement and gives suggestions for improvement. The objective of this study is to determine the optimal review time such that there is enough time to make judgments about changes required before the scheduled release. Design/methodology/approach Testing utilizes majority of time and resources, assures reliability and plays a critical role in release and warranty decision-making reviews necessary. A very early review during testing may not give useful information for analyzing or improving project performance, and a very late review may delay product delivery and lead to opportunity loss for developers. Therefore, it is assumed that the optimal time for review is in the later stage of testing when the fault removal rate starts to decline. The expression for this time point is determined using the S-curve 2-D software reliability growth model (SRGM). Findings The methodology has been illustrated using the real-life fault datasets of Tandem computers and radar systems resulting in optimal review time of 14 weeks and 26 months, respectively, which is neither very early in testing nor very near to the scheduled release. The developer can make changes (more resources or postpone release) to expedite the process. Originality/value Most of the literature studies focus on determination of optimal testing or release time to achieve considerable reliability within the budget, but in this study, the authors determine the optimal review time during testing using SRGM to ensure the considerable reliability at release.

Structural Health Monitoring (SHM) of large structures is a critical aspect due to various environmental conditions, high speed & long-distance communication, dynamic analysis of the structure, and cost of operation. These issues can be addressed using Fiber Bragg Grating (FBG) sensor technology which has evolved to a new height and is widely used in various distributed critical sensing applications. These are mostly preferred due to long-distance monitoring, low cost of operation, and immunity to Electromagnetic (EM) radiations. Similarly, the monitoring of a large structure from a long distance is also one of the crucial aspects of SHM technologies. These technological challenges can be addressed using an integrated distributed sensing solution consisting of FBG sensors, Big Data, Kafka, and the Internet of Things (IoT). In this article, the fabrication of the FBG sensor and the bonding of the sensing element to the base plate of the suspension bridge structure are discussed along with experimental details. A scalable architecture of the proposed Smart Distributed Sensing (SDS) model using FBG sensors is also discussed in this article. The experimental validation is performed using an IoT based FBG sensing mechanism to estimate the strain distribution profile at the bonding region of the base plate from a central location.

As the energy demands of the world are increasing, researchers are in search of those energy resources which are economical, easily accessible, versatile, easy to transport and produce very less residue and pollution on energy generation. Oil has been the world’s major commercial energy source from many decades and is likely to dominate in the twenty-first century as well. Oil accounts for nearly 40 percent of world’s energy mix. The aim of this research is to focus on availability of proved oil reserves along with production and consumption of oil in different regions and countries of the world. An effort has also been made to find the global position of India in terms of proved oil reserves, production and consumption of oil in the world. At the end, life of available proved oil reserves in the world is found out at the production rate of oil as in 2018. The management policy that may be adopted to use oil in a better way, considering environmental pollution, are also discussed.

Reliability of software products can be affected by several factors faced during the testing phase such as the generation of additional faults during the fault removal process, inefficiency or inaccuracy of team in completely removing faults, probability by which faults are removed, and difference between the number of failures and faults removed. These conditions have been termed as error generation, fault removal efficiency (FRE), imperfect debugging parameter, and fault reduction factor (FRF) respectively. In this paper, we have proposed a unified software reliability growth model (SRGM) to assess the impact of these parameters on the reliability and release schedule software. The error generation, imperfect debugging, and FRE parameters have been assumed to be constant while FRF is time-dependent modeled by exponential, Weibull, and delayed s-shaped distribution functions. These models can be represented using a single unified SRGM that reduces the difficulty of model selection. The special cases have been validated on two real-life fault datasets of Tandem computers and radar systems. Performance measures, the goodness of fit, and boxplot analysis show that the model fits the dataset very well. Further, the model that gives the best fit has been used for release planning for determining release and warranty time that minimizes development cost. The paper also analyses the sensitivity of cost parameters, the impact of change in error generation, FRE, FRF, and the reliability constraint on optimal timings. The results are highly encouraging for software managers and engineers and add value to the existing literature.

The software systems prevail various types of bugs reported by users. On the basis of severity levels of bugs, it is decided which bug needs to be fixed first. If the bug is crucial, then it is fixed immediately, and if the bug is minor, then its fixing procedure can be postponed. This research seeks to present an automated bug severity prediction model by using topic modelling and classification techniques. The summary of historical bug reports of the Eclipse project is pre-processed to extract features. Later feature extraction is done by using term frequency and latent Dirichlet allocation (LDA): a topic modelling technique. The topics extracted from LDA are then used to train the classifiers. Our approach also endorses the severity level for the newly disclosed bug. After training the model, the results actively exhibit admissible accuracy scores.

Link prediction is a widely studied topic in graph data analytics and finds numerous applications like friend recommendations in social networks and product recommendations in e-commerce. It refers to predicting new connections or edges that may arise in the near future amongst the nodes of the network. There exist various methods of link prediction generally based on either local, semi-local, or global features of networks and usually suffers from the problems of consistency in their performances over different and large size networks. In this paper, we intend to propose a generic and improved method of link prediction named as NSMLLP by integrating Node centralities, Similarity measures, and Machine Learning classifiers. We calculate popularity measures for every node and evaluate similarity measures for every pair of nodes in the network. The combined popularity and similarity measures form the features for every node pair in the network. The combined features of the nodes at the end of the edges, along with the positive or negative edge label, form a well-defined dataset for the task of link prediction. This dataset is then fed into machine learning classifiers like Random Forest classifier, AdaBoost classifier, and an ANN based classifier. The results obtained from these classifiers are then combined to make the final link prediction. We provide an information gain study aiming to quantify the improvement brought on by our proposed method. A feature importance study is also provided to comprehend better the relative importance of the various popularity and similarity measures used by us. The experimental results obtained on multiple real-life networks demonstrate that the proposed technique outperforms many existing popular methods of link predication based on several evaluation criteria.

The open-source software (OSS) market comprises of thousands of products and applications with different quality. The primary concern of the individuals and organizations is to evaluate the quality OSS products and packages. In this chapter, we have demonstrated a MCDM based model to assess the quality of OSS by taking performance and cost-based criteria related with OSS to analyze its quality on the basis of feedback gathered from the users and the experts. To avoid the uncertainty attached with the opinion of the expert, the Maximum-Entropy-Minimum-Variance-Ordered-Weighted-Aggregation (MEMV-OWA) operator has been incorporated. The criteria weights are calculated by solving a non-linear multi-objective programming problem proposed in the MEMV-OWA operator. The picture fuzzy set information has been used which is an addendum of intuitionistic fuzzy set, representing the human opinion more precisely. In this chapter, we have proposed a model named MEMV-OWA-PF-TOPSIS (Maximum Entropy Minimum Variance-Ordered Weighted Aggregation-Picture-Fuzzy-TOPSIS). A step by step procedure has been exhibited to show its implementation in real-life problems. A numerical illustration related to the software quality assessment of OSS on the basis of their performance and cost-based criteria has also been provided in this study.