The demand for explainable energy forecasting models has been emphasized in recent studies to enhance decision-making transparency. The efficient conversion of solar energy is stochastic and depends on environmental conditions. To address this, we propose a model-chain framework named XPEMM (XAI-based Photovoltaic Energy Management Model chain) for Solar Energy Production and Management in Smart Cities, ensuring explainability using LIME. Our model chain consists of three stages: Long-term forecasting of Global Horizontal Irradiance (GHI) as the first stage, Regression Modeling of GHI with power as the second stage, and Model-Agnostic explanation generation as the third stage. A recursive forecasting strategy was employed, leveraging Recurrent Neural Networks (RNN) to reduce error accumulation over a 5-year forecast horizon. The output of the proposed model chain assists grid cell operators in real-time monitoring, resulting in optimized performance by predicting the causes of high and low performance in advance, allowing for necessary adjustments. It also aids in site selection for high-budget PV-plant installations. We used two different datasets for training and benchmarking the Recursive Multistep GHI Forecasting Model (RMGFM). The proposed recursive forecasting strategy has been evaluated using RMSE and MAE error metrics which are 91.40 and 47.32 respectively, which are 2.9% 29.39%, and 7.9% less than the previously established approach using LSTM, PCR, and SVR based on RMSE measure and 15.9%, 49.9%, 34.8% less based on MAE metric. This study concludes that a recursive approach is best for long-term forecasting of GHI in regions with cyclic climatic patterns based on RMSE and MAE values. The interpretation of the LIME output establishes the fitness of model agnosticity in the Photovoltaic model chain.

Agriculture is the backbone of every country. The growth rate of the population is exponential and uncontrollable. Meeting the needs of the whole population is a strenuous task in the case of traditional agriculture. Precision agriculture is helping in attaining this goal by improvising the farming techniques to get more yields of crops. Precision agriculture is the process of embedding or using information technology in the process of agriculture to help monitor and analyze the crops. For this, cyber-physical systems (CPSs) are helping toward attaining precision agriculture. CPS is the way of integrating computation, networking, and physical processes. Nowadays, agriculture needs data on its elements to produce a better yield. Thus agricultural CPSs come into play. Thus, in this chapter, a discussion on how different information technologies have converged to bring out a superior form of agriculture with the usage of blockchain, 5G, Internet of things, cloud computing, ML, and big data.

Extended Virtual Reality has expanded its wings to almost each and every sector enabling immersive experience in various fields and has found applications in gamification, learning, healthcare, etc. This technology has aided in providing solutions to various problems in different fields, and healthcare is the most prominent one among them. Children suffering from ASD which is a developmental disorder affecting the brain that impacts how a person perceives external responses, are finding it increasingly difficult to get treated as the treatment methods are tedious. There are very few methods which are regarded as standardized means of treating autistic children but there are a few common traits that can be found in children affected by ASD which can be grouped under three common categories. They are lack of communication skills, lack of basic mathematical knowledge and low levels of remembrance. With the help of Gamification, which provides therapy by means of games to those affected, the kids affected by ASD can be treated, powered by the concept of Extended Virtual Reality. In this paper, we have developed a model to provide autistic children a real world experience of playing games which will help them in enhancing their skills without any external interferences. Children who play these Extended Virtual Reality based games show gradual improvement, for which the results can be facilitated with the help of a Linear Regression model, helping us predict future response times. The proposed model results in enhancement of memory levels of the kids as a result of the game and classifies kids based on their enhancement in memory into high, medium and low. The mean absolute error of the linear regression model is found to be 0.0394.

Throughout the last decades, the number of vehicles on the road has steadily increased due to the rising demand for urban mobility and contemporary logistics. Two of the many detrimental effects of more vehicles on the road, which also impede economic development, are increased traffic congestion and traffic accidents. The issues mentioned above can be significantly resolved by making vehicles smarter by reducing their reliance on humans. Over the past century, various nations have conducted extensive research that has fueled the automation of road vehicles. The development of autonomous vehicle (AV) technologies is currently being pursued by all significant motor manufacturers worldwide. Undoubtedly, the widespread use of autonomous cars is more imminent than we realize given the development of artificial intelligence (AI). In order for AVs to perceive their surroundings and make the right decisions in real time, AI has emerged as a crucial component. This development of AI is being driven by the growth of big data from numerous sensing devices and cutting-edge computing resources. We must first examine AI's development and history in order to comprehend its functions in AV systems.

The potential for connected automated vehicles is multifaceted, and automated advancement deals with more of Internet of Things (IoTs) development enabling artificial intelligence (AI). Early advancements in engineering, electronics, and many other fields have inspired AI. There are several proposals of technologies used in automated vehicles. Automated vehicles contribute greatly toward traffic optimization and casualty reduction. In studying vehicle autonomy, there are two categories of development available: high-level system integrations like new-energy vehicles and intelligent transportation systems and the other involves backward subsystem advancement like sensor and information processing systems. The Advanced Driver Assistance System shows results that meet the expectations of real-world problems in vehicle autonomy. Situational intelligence that collects enormous amounts of data is considered for high-definition creation of city maps, land surveying, and quality checking of roads as well. The infotainment system of the transport covers the driver's gesture recognition, language transaction, and perception of the surroundings with the assistance of a camera, Light Detection and Ranging (LiDAR), and Radio Detection And Ranging (RADAR) along with localization of the objects in the scene. This chapter discusses the history of autonomous vehicles (AV), trending research areas of artificial intelligence technology in AV, state-of-the-art datasets used for AV research, and several Machine Learning (ML)/Deep Learning (DL) algorithms constituting the functioning of AV as a system, concluding with the challenges and opportunities of AI in AV.

Artificial intelligence is now a necessary component for both production and service systems in recent years, as technology has become a vital aspect of daily life. Automated driving vehicles operate autonomously, also known as driverless cars that can operate without a human driver. Research on autonomous vehicles has substantially advanced in recent years. Artificially intelligent autonomous vehicles are the current need of the society. Although some people might be apprehensive to give a computer control of their vehicle, automated driving technologies have the potential to make roads safer. Self-driving automobiles can address environmental issues as well as safety-related ones. Unlike humans, computers do not really have difficulty keeping attention when driving. Additionally, by responding appropriately, an automated car can prevent accidents to potentially dangerous events on the road. Self-driving technology has many advantages, one of which will make more easily accessible means of transport to people who are unable to drive. For a variety of reasons, such as inexperience, incapacity, or age, many people are unable to operate a vehicle. These individuals can travel considerably more safely and independently. Therefore, we will explore the architectures of both software and hardware of autonomous cars in this chapter, as well as their parts, benefits, and future developments.

The subset of Artificial Intelligence is Deep Learning which is inspired from the arrangement and communication of neurons in the Brain. 2D-CNN models are used to propose a solution to the Separation Problem. Various previously trained Architecture in large databases such as VGG-16, V66-19, Inception V3, ResNet-50, DenseNet-201, etc. are available, which can be used, this technique being called as Transfer learning. The process of finding an object in a picture is called Object Detection. There may be more than one event of the same object or more than one type of object in the same image. Many of the initially proposed solutions to this problem depend entirely on the first proposed districts for procurement. Among the many proposed region proposal Algorithms, Complete Search, Selected Search, Slide Window and Edge Boxes are some of the algorithms most commonly used for object discovery. R-CNN, Fast R-CNN, Faster R-CNN, Mask R-CNN, YOLO, SSD, and Detectron can be said to be the most advanced and extensively used approaches, each with its own drawbacks and advantages of Object Detection. Splitting is the work of combining image pixels together based on a specific principle. The Mask R-CNN Algorithm also deals with the problem of fragmentation. This chapter is designed to elaborate on YOLO Object Detection Algorithms and SSD, a comprehensive application of Object Detection Algorithms in Clinical Image Analysis, and the use of YOLO Architecture to detect brain tumors from MRI Image results.

Detecting road signs is a critical element in cars and driver support systems. This system is separated into two parts: road sign detection and categorization of the recognized sign. This system is constructed with a CNN, which can recognise traffic signs in photos and categorise the image into a certain class to which the sign belongs. CNN is utilised because it is the best for picture categorization. First, the photos are transformed to greyscale and the parameters are optimised. The margins of the traffic signs are tightened in the following phase of the procedure. Following that, the area of interest is gathered and input into the convolution neural networks. CNN then classified the images into their classes and with an accuracy of 99.62 percentage.

Human Activity recognition is an active field of research in the current days with more and more advancements in the IoT technology introducing the culture of gadgetization of electronic goods in the world. Smartphones-based HAR is an efficient way to monitor human activity without the need of any specifically designed assistive tools for applications in healthcare monitoring, patient monitoring, management of elderly people etc. Deep learning algorithms reported better performance in comparison with machine learning algorithms, however they are criticized in aspects of computational requirements. Resource constrained Edge-based implementations demand for the design of light-weight models. Hence, we propose a 2-step HAR pipeline involving Random Sampling-based data pre-processing. The proposed approach uses relatively less data for training and the final model based on convolutional neural networks is light-weight involving a low number of model parameters and achieves an accuracy of 96.5

The term ‘Fashion technology’ is used to refer to innovative methods of creating new designs for clothes and also sourcing them. It facilitates the formation of new tools for the fashion industry which increases the manufacturing and utilization the fashion products. In this era of fashion, manufacturing processes, shipping and the selling of fashion products make use of the fashion design technologies available to the fullest. As of now, the growth rate of fashion industry is very high. The fashion space is being automated, personalized and sped up by a list of inventions which include the clothes that can be worn with the help of virtual reality, algorithms in AI which can predict the trends that follow, and Robots which cut fabric. Advanced technologies like blockchain and virtual reality are adapted by many fashion brands and companies to create a unique purchasing experience for the customer. Fashion technology cannot escape the fact that innovations help industries to evolve and adapt to the existing environment faster and that it is no exception. A lot of effort has been made to customize the clothes of users in order to help them wear those matching their exact body dimensions. This paper aims to customize purchasing the clothes by using 3D body scanners to get the exact measurements so that the user can wear readymade clothes according to his exact body dimensions.

The availability of labelled data at scale has contributed to the rapid expansion of Deep Learning. This results show that training deep-cnn learning representations can lead to huge increases in performance on specific tasks. The caffe net model surpasses existing modernization approaches for gender and age approximation when tested against the Adience benchmark. It is suggested convolutional net architecture is simple and accurate to 95%.

Abstract Sustainable Energy alternatives such as Solar Energy, Wind Energy are the best alternatives to the harmful Non-Renewable energy sources like fossil fuels. With the increasing investment made in the Sustainable Energy alternatives, forecasting of Energy production from the farms plays a crucial in the structured design of Energy farms and for integrated smart grid management. In this research, we propose SRNN-LSTM hybrid models for univariate solar irradiance forecasting, multivariate temperature forecasting and univariate forecasting of Wind Speed. The proposed models are evaluated based on RMSE, MAE, r2 score and Index of Agreement error metrics. The SRNN16LSTM16 hybrid models achieves the best performance with r2 score of 0.97 for Solar Irradiance forecasting and SRNN8LSTM8 hybrid model achieves the best performance with the r2 Score of 0.989 and RMSE of 0.122 for Wind Speed forecasting. We have also proposed a simple approach to derive the prediction intervals based on the Median of the residual values. The higher values of PICP error metric indicate that the proposed models are efficient and perform better than the state-of-the-art models used in the forecasting frameworks. We have also performed the Model Analysis using Explainable Artificial Intelligence tools such as LIME and ELI5 to analyze the importance of features in the model development process, generating and analyzing local explanations.

In the pharmaceutical industry, accurate demand forecasting is crucial for the efficient management of manufacturing, acquiring, and distribution activities. This paper concentrates on utilising data mining techniques, namely the Decision Tree and Random Forest algorithms, to forecast the demand for pharmaceuticals associated with seasonal ailments that are anticipated to emerge in the upcoming months of 2024. The findings of the analysis demonstrate that the Random Forest algorithm exhibits superior performance in comparison to the Decision Tree algorithm. This is evidenced by the observation of lower mean Root Mean Square Error (RMSE) values of 80.53 and 97.10, respectively, which indicates that Random Forest outperforms Decision Tree. The present work's results offer significant contributions to the pharmaceutical sector by providing valuable insights that facilitate improved planning and resource allocation to address the variable demand patterns that are influenced by seasonal diseases.

Deep Learning has been a revolutionary innovation in the field of medical imaging. The domains that once required hours of intense study for detection or classification of a disease has now exponentially reduced with the help of certain state of the art works of DL. In this work, we have proposed two types of classification procedures for the diseases Miosis and Mydriasis which unlike Anisocoria, extremely dilates or constricts both the pupils. This condition is popular among people with brain disease, traumatic brain injury and by medications like opioids. This is also common in the field of agriculture as one of the causes being direct eye contact with chemicals such as pesticides. The proposed approaches are based on Convolutional Neural Network and Hough transformation techniques for identifying the arbitrary shapes of iris and pupil.

In the past two decades, there has been a sharp rise in the use of deep learning for medical image processing and analysis. Recent challenges, for instance, the most well-known ImageNet Computer Vision competition, have almost entirely incorporated deep learning approaches for providing the best result. The concept of Image classification was later extended to Image Segmentation and Object Detection which proved to perform extremely well using state-of-the-art classification algorithms as their backbone architecture. The accuracy of the algorithm and approach has a significant impact on the medical field as there is a constant need for accurate and computationally efficient models. The existing object detection and segmentation approaches need large data for providing accurate results, unlike classification algorithms in which accuracy can be achieved with a relatively smaller amount of data. Hence, for the overall increase of model accuracy, there is a need for image augmentation to be incorporated. In this paper, several deep learning methodologies such as classification, object detection, ensemble, and segmentation for pneumonia classification and detection have been reviewed and an ensemble-based approach for the classification of Pneumonia using chest X-rays has been proposed.

The neurological ailment Parkinson's disease affects millions of individuals globally. Contrarily, an early detection of the illness will aid in its efficient treatment. With machine learning, which is still in its infancy, there are a number of prospects for computer-aided identification and diagnosis that may lessen inescapable medical issues and intrinsic clinical uncertainty, offer direction, and enhance decision-making. In an effort to improve the technique for diagnosing Parkinson's disease, we have proposed a model for early detection of the condition. Performance and speed were taken into consideration when XGBoost, a new machine learning technique, was developed. A decision tree-based technique called Extreme Gradient Boosting, or XGBoost, was developed.