VOLUME 9, ISSUE 12, DECEMBER 2020

Abstract: This paper presents the development of an electrical model for an implantable electrode that allows the detection of the most stable and precise electromyography signal (EMG) directly on a skeletal muscle in a chronic way.

Keywords: Implantable epimysial electrode, electrical activity of a skeletal muscle, electromyography (EMG), electrical model of an implantable electrode

Abstract: Gesture language recognition is a technology that allows humans and animals to interact with computers and hardware without the need for any mechanical actuation between them. Gestures can include anything from facial expressions, head or eye movements, hand and arm motion, or other bodily movements and can therefore encompass a wide range of control for complicated systems. As sensor technology becomes smaller, cheaper and more readily available, gesture language recognition becomes a more viable solution for projects and systems to employ. This paper presents a prototype gesture language recognition system composed of hobby-grade electronics, compares it against systems using higher-grade hardware or more advanced software algorithms and extrapolates what further technological advancement will do for gesture language recognition systems.

Keywords: Gesture Language, MicroElectroMechanical Systems, Inertial Measurement Unit

Abstract: Cloud computing is advanced technology, provides a road map to access the applications over the net. Cloud client and information owner customise applications through web. Because of storing Brobdingnagian quantity of knowledge on cloud, there could also be several problems associated with the protection in cloud network. To describes a privacy conserving cluster ways, homomorphic cryptography schemes which will run on a typical high performance computation platform, like a cloud system. In existing system offers a privacy conserving distance matrix calculation for many cluster algorithms. The privacy model is to applying part homomorphic cryptography ways to make a probabilistic classifier victimisation the acute learning machine rule and created the privacy-protected version of the ELM rule, that constructs a classification model by making an equation. In proposed model could be a privacy-preserving methodology victimisation the Paillier cryptography system for outsourced sensitive datasets. The consumer builds a final cluster model with aggregation of every encrypted distance matrix calculated at each party. Additionally work, so as to forestall information speech act from the model, the model itself ought to even be encrypted victimisation homomorphic cryptography algorithms. To permit the consumer to use the encrypted cluster model, new K-HUB cluster models square measure developed.

Keywords: Clustering; Homomorphic cryptography; Machine learning; Paillier encryption.

Abstract: Sleep is essential for humans although its basic physiological function remains obscure. Humans suffer from various sleep disorders including dyssomnias such as insomnia, hypersomnia, narcolepsy, parasomnias, bruxism, circadian rhythm sleep disorders and sleep apnea. The major problem addressed in this review paper is Obstructive Sleep Apnea in geriatric population. Sleep Apnea is defined as hinderance of airflow during breathing where soft tissue muscle in throat relax and narrow the airway which cause blockage of upper airway tract (Obstructive sleep apnea). Microcontroller based sleep apnea monitor consists of a Respiration sensor for measuring and monitoring the breath condition, SPO2 and Pulse Sensor for monitoring the heart rate and blood oxygen saturation level and Digital Humidity and Temperature sensor. All the parameters are monitored by the microcontroller once the microcontroller detects any difficulties in breathing conditions it’ll automatically turn on the supportive system for the patient. By using our proposed supportive system patients can able to breathe stably. In addition, the acquired parameters will be accessible to the practitioner through customised IoT platform which is inbuilt in system.

Keywords: Dyssomnia, Obstructive sleep Apnea, Geriatric population, Microcontroller and IoT

Abstract: Smartphones are revolutionary mobile devices, can be found with almost every person of all age groups, become an integral part of everyday life. Use of the smart mobile devices increases by leaps and bound with the growth of location-based services (LBS). With the ease of availability of internet and GPS enabled mobile devices with high accuracy and precision, location-based queries become very popular among mobile users on the road network. In this paper we have intensively surveyed cutting-edge solutions approaches for types of spatial query, mainly, location-based proximity queries (nearest neighbour queries, k-nearest neighbour queries, and range queries) on the real road network.

Keywords: Proximity query, Spatial road network, Location-based services, Range query, Nearest Neighbour query, k Nearest Neighbour query.

Abstract: This paper presents the design of an Infinite movement of a circular wooden wheel from iron ball. So it is entitled as endless energy using rotational movement with its own input. With an objective to continuously movement of the wheel, we use wooden wheel with four holes at the side of the wheel and a hole at the centre of the wheel along with four iron balls, spring, iron rods, angular rod, gear and bearing. The design is discussed in details later in this paper. There are various benefits from the design.

1. a) There may be endless energy when the continuous movement of the wheel from the iron ball strikes at the corner of the wheel.
2. b) It may help to generate electricity as an output when gear is connected at the centre of the wheel.

Keywords: endless energy, electricity, iron balls, spring and wooden wheel.

Abstract: (Naya-Plasencia et al., 2010)The ESSENCE group of CHFs, planned by Martin, was a Round 1 competitor in the SHA-3 rivalry. It is a group of square code based CHFs utilizing the Merkle Damg'ard method of activity. The ESSENCE family employments basic calculations that are effectively parallelizable and entrenched numerical standards. Quintessence accompanies a proof of protection from direct and differential cryptanalysis that until this work stayed unchallenged. In this paper, we first depict a few undesired properties of the ESSENCE L capacity. These can be utilized to fabricate a sans semi beginning crash assault on 31 out of 32 rounds of the ESSENCE-512 pressure work utilizing a differential trademark. (Knopf, 2007; Naya-Plasencia et al., 2010)This straightforwardly refutes the plan guarantee that 24 rounds of ESSENCE make it impervious to differential cryptanalysis. To fabricate our assault, we depict a novel procedure to fulfill the conditions forced by the trademark in the first nine rounds. We don't know about a comparative procedure in existing writing.

Abstract: Digital color images are the most important data types used in various vital applications, some of these applications required image features. Digital images contain variable number of objects with variable sizes. Each object contains valuable information which can be used to construct image features. In this paper research we will analyze a variety of information which are used to describe any object in the image, a methodology of objects extraction and the associated with each object will be proposed, a way of forming the image features will be discussed. Key words: Digital image, object, features, centroid, area, extrema, orientation, convex Hull, Euclidian distance.

Abstract: With the expansion in the interest for superior and fast VLSI frameworks, such as network Processors In networking or SOCs in communication and computing has shifted the focus from traditional performance Parameters towards the number of chips, LUT’s etc. and frequency consumption. Paired tree topology s the one of the topology for system on chip plan in this proposition we have effectively structure the equipment chip for bunch size 2, 4, 8 and 16 separately. The capacity on recreation is done in Modalism programming the outcomes are checked on Spartan 3EFPGA. Our binary chip is optimization in terms of hardware parameters such as LUTand flip- flop. The system supports high frequency.

Keywords: VLSI, SOCs, FPGA, LUTs.

Abstract: Security is a challenge in every secured areas and devices access. IoT devices with modern AI are using in access control modules to permit it verified authority. In this paper we consider user experience from their daily uses of IoT devices. Finger print based access control module handle images from users with seasonal skin issues, pressure of fingers, placement angels of fingers and carelessness of users. Existing IoT devices access control module domains highly concern on better matching of pattern. Though person is verified, some issue arises to access control that generate negative mind to users of IoT devices. Institutions maintains manual register book to support the crux of IoT devices access control failure. We design a repeated partial and adaptive model to overcome the problems of existing models. This approach will help to assure better access control in IoT devices and users experience. IoT and AI industries will be highly beneficiary from this research to assure quality production.

Keywords: AI, IoT, Access Control, Pattern Matching

Abstract: Power of sight is considered as the most important of all senses. Blind people are often dependent on others for their daily tour. Over time several advancements in technology has helped in increasing entertainment and comfort for the blind. “Smart shoe" can assist the blind on their daily routine and can act as a comfortable and safe companion on their journey. Common assistances provided earlier for the blind include walking sticks or guide dogs. It includes ultrasonic sensors with a step counter that can alert the person of the impending obstacles. The technology results can improve their ways of growth and can drive them to lead their independent lives.

Keywords: Visually impaired, Navigation, Step counter, Sensors, obstacles

Abstract: Lots of Applications are developed to track a Smart phone but still it is a major concern. User has to manually report to the customer care to block the number of the lost Phone. So that, Android Application is deployed with initial registration of Alternative Mobile number. An Application which is deployed in the mobile devices can be able to Track the current location of the device. If the robber changes the SIM card, immediately then location details are sent to the alternative Phone number of the original User. In this paper, both the logic of tracking the Theft Phone with mobile number & with GPS is tracked continuously. The registered alternate mobile numbers can get the SMS alert from the Theft Mobile. This process is reworked continuously to track the android mobile phone. Our Anti-theft mobile application will be capturing the picture of the thief as soon as the mobile is switched on and app launched and will be sending the location to the alternate mobile number. The location will be sent as a link of the map on which the location traced with longitude and latitude which can view in maps with marker to location. It will be needed an active internet connection to make the process possible, the user can register for it’s lost phone on our web application through which the mobile application will be triggered. Once the application is triggered with lost status, it will start working and as soon as the app launched, it will click a picture of the thief and will send it along with the location to the alternate mobile number mentioned during the registration process. This will help the user to trace his/her mobile phone without third party help.

Keywords: Location tracking, Android, Smart Phone, Context, Web Application, SMS Services, Picture Capturing

Abstract: Financial supply chain management (FSCM) is the practice of looking at all your financial processes at the holistic level, rather than viewing them as individual processes. It’s the end-to-end process that involves the procure-to-pay cycle, working capital management, and the order-to-cash cycle business processes. This field is a relatively new one. Despite the crisis-enhanced research interest and the growing importance of FSCM, academic contributions and discourse on the subject remain fragmented and vague. Mainly conceptual approach dominates; research methods employed are mostly empirical surveys and case studies, with the main focus given on the manufacturing industry while the bank “dimension” in the FSCM equation is neglected. At the same time, scarce research efforts have been identified towards the systematic documentation of the core concepts and formative elements of FSCM in the direction of building a “general theory of FSCM”. The paper provides a literature review of the FSCM discipline, identifies gaps and challenges in the field while providing insights on a future research agenda, thus preparing the ground for FSCM standardization and hopefully initiating a fruitful academic dialogue on the subject. A review and analysis of selected, up to date theoretical and empirical literature is provided in order to prove the significance of this discipline in modern management theory and provide useful conclusions. Moreover, an emphasis is given on the contemporary aspects of FSCM in terms of collaboration among companies, suppliers and financing institutions.

Keywords: Supply Chain Management (SCM), heated competition, Financial Supply Chain Management (FSCM), modern Information Technology, logistics, crisis, financial performance, Literature Survey and financial processes.

Abstract: As we all know that today’s life is very fast and the rider kick the and move forward without removing the side stand because of hurry and this may cause accidents. When accidents happen, it causes severe injuries and sometimes deaths as well. There are many reasons for these accidents, internally and externally. Forgetting to lift up the side stand is one of the reasons and it could be avoided by certain technical aspects. This paper presents an automatic side stand for motor bikes, which could be lifted up while the bike starts moving. The design consists of sensors, motor which are controlled by Arduino UNO program. There is no need of additional energy source or any other complexity since the designed system uses the necessary power from the motor bike battery.

Keywords: Human carelessness, solution,sensor,automation

Abstract: Agriculture is one of the main factors contributing to the economic growth of many developing nations. It is also the primary source of livelihood of many people in different parts of the world.The system designed and implemented here is a smart farming system which can handle almost all major facets related to irrigation and crop maintenance. From the farmer’s point of view, smart farming should provide the farmer with added value in the form of better decision making or more efficient exploitation operations and management.

Keywords: IOT, ponsing, Raspberry Pi, LCD, photoresistor

Abstract: Artificial Intelligence (AI) is knowledge exhibited by a machine rather than human insight. The field of fake wisdom depends on why human experience can be so decisively portrayed and reenacted by a machine. Proof of the positive effect of AI frameworks is surrounding us. It will never be feasible for such devices to supplant human creatures somewhat because AI comes up short on a human touch. Associations and people far and wide are making center standards around AI with an accentuation on a more humanist methodology. They understand the favorable critical circumstances AI can bring to their business. This paper gives a prologue to acculturated AI.

Keywords: Artificial Intelligence, Artificial Intelligence, Humanized AI, Human AI System.

Abstract: Digital audio files are one of the important types of data to be used in many vital applications, which leads us to protect this type of data from intruders and snoopers, as it may contain confidential information that not allowed to be viewed by any unauthorized third party . In this paper, we present a new method for protecting audio files that depends on the use of the digital color image or part of it as a key to the encryption or decryption process. The digital image used as a key is characterized by its enormous size, in addition to that, the image is not circulated through social media, but rather is agreed upon between the speaker and the listener, which leads to difficulty being known by the interlopers, and thus we provide a safe way to deal with audio files. The proposed method will be implemented and tested to measure the efficiency and quality factors.

Keywords: Speech file, encryption, decryption, WPT, YIQ, MSE, PSNR.

Abstract: In this project, we use a completely deep learning based approach to solve the problem of object detection in an end-to-end fashion. The network is trained on the most challenging publicly available dataset MS COCO like (SSD, RCNN, Faster RCNN, YOLO v3, 4 etc.), on which object detection challenge is conducted annually. The objects are detected in boxes by this dataset where objects like car, bike, person, etc.

Keywords: Object detection, convolution neural network, scoring system, selective search, deep learning, MS COCO, SSD, RCNN, YOLO, OD model.

Abstract: a single phase grid-connected converter is usually adopted In low power renewable distributed systems. This paper deals with a A review of the multi level topologies , a theoretical power loss comparison with the proposed solution is realized. The proposed converter is full-bridge architecture with two extra power switches and to the midpoint of the dc link two diodes are connected . Since the two more levels are obtained when two capacitor of dc link is discharge, the balancing of the midpoint voltage is achieved with pulse width modulation (PWM) strategy, Multilevel converters have been under research and development for found successful industrial application. This is still a technology under development and many new commercial topologies have been reported in the last few years advances made in modulation and control of multilevel converters are also mention. A great part of this paper is to show non traditional applications powered by multilevel converters and how multilevel converters are becoming an enabling technology in more industrial areas.This technology developed for renewable energy scheme where unity power factor required . a variation of the proposed topology which allows four-quadrant operations.

Keywords: Cascaded Full-Bridge, Hybrid Five-Level Topologies, SPWM, THD, DC to AC Conversion Distributed power Generation, grid- connected converters, Single – phase system multilevel converters. Multilevel power conversion; Power quality; Harmonic reduction

Abstract: The main aim of this virtual placement portal is to unify the on-campus placement process across various institutes i.e., providing a common platform where all registered colleges can conduct the on-campus placement process virtually. Having moved to complete “virtual” placements for the 2020 batch due to the COVID-19 situation, some of the key issues were observed and hence to address such issues the online placement portal is being developed as solution that not only tackles the noted issues but also bridges the gap between the placement cell of an institute and its students. The portal helps keep the students updated about all companies that visit campuses for placements, directly by the placement cell of the institute with little to no intervention of any intermediary link between the students and the placement cell such as the "placement coordinators".

Keywords: Virtual recruitment, Web portal, Placement- management system, Web-based system

Abstract: Fraud detection remains an active area for research. While fraud is a difficult problem hampered by issues ranging from uncertain data to adversarial environments, new technologies and techniques from the fields of data science, machine learning, and big data bring opportunities to alleviate some of the difficulties. In this paper, fraud detection is examined in the context of financial systems for public-sector applications. Public sector financial systems face challenges with fraud detection, such as a large volume of transactions and a need for complex monitoring rules based on the context of transactions. The analysis of public sector financial systems is framed as a data-driven approach to understand the domain. Initial steps to facilitate the analysis of data within a public-sector context are taken by proposing a model framework consisting of detectors, monitors, and pattern mining techniques along with an input data requirements and output results taxonomy. The frameworks’ components are investigated and elaborated on in terms of the public sector financial systems. Furthermore, future directions toward further development and evaluation of the components within the context of public sector financial systems. Fraud detection remains an active area for research. While fraud is a difficult problem hampered by issues ranging from uncertain data to adversarial environments, new technologies and techniques from the fields of data science, machine learning, and big data bring opportunities to alleviate some of the difficulties. In this paper, fraud detection is examined in the context of financial systems for public sector applications. Public sector financial systems face challenges with fraud detection, such as a large volume of transactions and a need for complex monitoring rules based on the context of transactions. The analysis of public sector financial systems is framed as a data-driven approach to understand the domain. Initial steps to facilitate the analysis of data within a public sector context are taken by proposing a model framework consisting of detectors, monitors, and pattern mining techniques, along with an input data requirements and output results taxonomy. The frameworks’ components are investigated and elaborated on in terms of public sector financial systems.

Keywords: Anomaly Detection,Predictive Analytics,Behavioral Modeling,Supervised Learning,Unsupervised Learning,Real-time Monitoring,Data Integration,Risk Scoring,Entity Resolution,Natural Language Processing (NLP),Data Lake Architecture,Feature Engineering,Graph Analytics,Model Explainability (XAI),Regulatory Compliance Analytics.

Abstract: In the dynamic landscape of retail manufacturing, effective demand forecasting hinges on sophisticated data engineering practices that streamline and optimize data flows. This paper explores the intricate processes involved in developing an end-to-end data engineering solution tailored for demand forecasting within retail manufacturing ecosystems. By integrating heterogeneous data sources, constructing robust data pipelines, and implementing advanced analytics, businesses can transform raw data into actionable insights that drive strategic decision-making. Leveraging cutting-edge technologies such as cloud computing, machine learning algorithms, and real-time data processing, this approach addresses the inherently volatile nature of consumer demand. The framework delineated in this work emphasizes scalability and flexibility, essential for adapting to the ever-evolving market conditions. Key components include data ingestion, cleansing, enrichment, storage, and the deployment of predictive models, each of which plays a pivotal role in refining data utility and forecasting accuracy. Particular attention is given to data governance and security, ensuring compliance with regulatory standards and fortifying data integrity. By adopting this comprehensive methodology, organizations can enhance their agility, mitigating risks associated with demand surges and supply chain disruptions. This study contributes to the discourse on integrating technological advancements in data engineering with practical demand forecasting applications. It demonstrates how leveraging data-driven strategies can optimize inventory management, reduce waste, and improve customer satisfaction. As retail manufacturing ecosystems become increasingly complex, the insights presented provide a blueprint for harnessing the full potential of data engineering, fostering innovation and competitiveness in the industry.

Keywords: End-to-end data pipelines, Retail demand forecasting, Data engineering for manufacturing, Predictive analytics retail supply chain, ETL for demand forecasting, Time series forecasting pipelines, Big data in retail manufacturing, Data lakes for supply chain analytics, Machine learning demand prediction, Real-time data integration, Forecast accuracy optimization, Feature engineering for demand models, Scalable data infrastructure, Cloud-native data platforms, Demand sensing with AI.

Abstract: Digital finance encompasses digital information feedback on credit credibility and electronic transactions based on digital currencies. In recent years, driven by blockchain technology and cloud computing, digital finance has boomed in developed countries, giving rise to new issues and potential risks. With the transformation of digital finance from 'crypto' to 'cloud', a new trend is emerging. Digital finance driven by cloud computing, whether on public clouds or on-chain clouds, offers a plethora of opportunities, but the migration to native digital finance systems based on cloud infrastructure would be fraught with risk. Cloud computing, with its convenience, flexibility, security and scalability, continues to become a mainstream technology in retail finance. Multi-modal federated machine learning and federated cloud transactions can potentially construct a decentralized market with permissioned access while achieving snapshot privacy. Blockchain will continue its widespread adoption in developer-centric financial backends as technology matures, with larger public chains evolving to handle benchmark trading volumes. On miniambilest, native public and enterprise banks, decentralized exchanges, NFT-backed loans and on-chain credit scores will rise, creating liquidity for illiquid assets and granting borrowing access to previously excluded participants. The rising popularity of cryptocurrencies may, on the contrary, accelerate KYC, AML and the adoption of CBDCs.

Abstract: Digital insurance services already provide consumers many benefits; however, they bring multiple risks as well. Privacy and personal data protection risks, attacks to the cybersecurity model, fraud, misinformation, among others, threaten these consumer benefits in a direct way. The growing challenge for policy makers is to guarantee data protection and control for consumers while not undermining the role of other stakeholders such as capital markets and financial institutions. Although the limited time frame for the implementation of the relevant regulations is a huge challenge for the relevant authorities, it should not lead to generic solutions, which without knowledge of the specific context and of the opposing stakeholder interests will not work appropriately. Moreover, they should not lead to models which cannot be implemented easily and quickly in practice. The design of transparent and resilient digital identification frameworks should take place after careful examination of the public and private ecosystems and of their intrinsic characteristics, in close cooperation with societal stakeholders, and focusing on proper incentive schemes. Insurance companies introduced improved scalability and resilience in their digital identity protection approaches, but at the resulting higher costs. The policy makers now face the challenge to improve the protection of personal data and identifications on the cloud infrastructure of the insurance companies. They should remain well aware of the fact however, that quick and easy to obey regulation seems impossible. For example, decentralized identity platforms are at an early stage. Consequently, their universal implementation cannot be quick, as many stakeholders should be engaged to successfully operate through decentralized identities. On the other hand, promoting some characteristics of decentralized identity solutions through standards (or at least usable solutions with such characteristics) could be done as an initial step in a quick and effective way. Finally, understanding of context and incentives for compliance should be introduced in the design of regulations at the outset. Privacy preserving solutions that protect the consumers data and identifications are not in the best financial interests of the insurance companies as there are important sunk costs in their current digital identity and personal data protection infrastructure. Understandably, their desire for a quick windfall gain in compliance will lead to a generic regulation finally hurting consumers rights.

Keywords: Identity and Access Management; Identity and Access Governance; Digital Insurance; Cloud-based Identity and Access Management.

Abstract: Cloud computing substantially reduces the carbon footprint of IT services, with consumption of energy from renew- able resources being the key driver behind energy-efficient design in cloud data centres. The energy footprint of applications hosted on public clouds is nevertheless non-negligible and sustainability- aware design choices should be employed to minimize it. Large- scale banking applications represent an ideal case for such patterns as they are expected to operate round-the-clock while generally being subject to low resource utilization for significant periods of time. Additionally, besides energy concerns, they are also subjected to nearby real-time performance SLAs, high avail- ability requirements, and predictable traffic provisioning. Typical energy-aware design choices made for services hosted on the AWS public cloud include using energy-demand-aware variants of the services (e.g. EC2 Reserved Instances instead of EC2 On-Demand Instances), reducing the traffic across regions, and adapting the allocation of network and compute resources (e.g. RDS read replicas) to the system load. The recommendations summarized here relate to a full-scale case-study banking application—the piece of project work—deployed into the AWS public cloud. The AWS support for energy-aware design patterns is illustrated by the practical application of System Architecture and AWS deployment design guidelines proposed for making the systems energy-efficient while meeting the aforementioned constraints. The suggested design patterns include principles for sustainable software architecture, patterns that drive power-aware design at the service layer, and techniques for optimizing compute and network resources under the joint action of energy demand and service responsiveness.

Keywords: Energy-Efficient Cloud Architecture, Green Computing in Financial Services, Sustainable AWS Deployment Patterns, Serverless Energy Optimization, Cloud Cost and Carbon Efficiency,E lastic Scaling for Power Reduction, Workload Scheduling and Auto-Scaling, Sustainable Fin- Tech Infrastructure, Carbon-Aware Cloud Orchestration, Energy- Optimized Microservices Design.

Abstract: This study presents evidence-based analysis of data-driven tax compliance monitoring within public revenue systems, emphasizing rigorous methodology, transparent reporting, and measurable outcomes. Emerging trends in the digitalization of the economy and public service delivery increasingly motivate the use of big data analytics by tax authorities for efficient revenue collection, including compliance monitoring. Political pressure or judicial rulings often push decision-making and applied analytics into unsafe territories. In contrast, peer-reviewed modelling and analysis support legitimate and transparent data-driven compliance monitoring. Digital economy taxation gap analytics is one area gaining widespread research and implementation attention. Another core aspect of tax compliance risk assessment is the detection of tacit and explicit noncompliance. Risk-assessment support for audit selection or the identification of high-risk sectors and segments is increasingly provided by data-analytical methods. While demonstrated in a financial-supply-chain context, the approach is transferable beyond the scope and context of any one implementation. Empirical examinations validate the proposed concepts, metrics, and models, establishing the methodological foundation for a broader application across public revenue systems. Supply-side data companies maintain data about every firm across digital supply chains. Cross-jurisdictional public-sector data-sharing silos constitute uncontested evidence of tax gap size and sector non-compliance risk or willingness summaries. Non-participation in regulatory audits, data-feedback mechanisms, and information-exchange strategies results in an ever-increasing compliance gap. Tax administrations and regulators have decision responsibility and operational authority for compliance governance, yet any failure to embrace any one of the risk-and-remediation strategies should also be held to account.

Keywords: Data; Tax compliance monitoring; Public revenue systems; Decision support; Policy evaluation; Policy design; Administration arrangements; Policy instruments; Tax gaps; Revenue risk.

Abstract: Smart agriculture innovation enables the intelligent optimisation of crop growth conditions based on environmental and soil parameters. A Cloud-Based Big Data analytics framework can be adopted to analyse different ingested data sources in order to develop models of predictive analytics for crop stages and assist the decision-making processes of farmers. Big Data Analytics entails several sophisticated methods and tools, which can be provided as a service using several services integrated under a Big Data cloud framework. The aim is to support the decision-making processes of farmers by providing a comprehensive view of the past, present, and future of farms. Data from diverse sources of different types and characteristics are injected into the system for cleaning and pre-processing. Big Data Analytics techniques are efficiently applied for predicting crop growth stages and diseases using predictive analytics, remote sensing, and computer vision technologies. Smart agriculture solutions should also consider the issues of data quality, privacy, and security. The continuous growth of the world population raises the need to increase food production. Agriculture and rural development must, therefore, remain top priorities for governments. As the population increases, the demand for food, clean water, and energy increases as well, and the challenge is to fulfil this demand. On the one hand, the response to this growing demand requires an evolution of the agricultural sector through the adoption of new technologies. On the other hand, climate change imposes a new set of challenges to farmers. In this context, information and communication technologies (ICTs) can help farmers increase productivity, fertilisation efficiency, irrigation application, and pest control while reducing operational and management costs. The proper combination of these technologies leads to Smart Agriculture.

Keywords: Cloud computing; agriculture; big data; Internet of Things; data analytics; data processing; machine learning; cloud storage; data acquisition; wireless sensor network; Apache Spark; fog computing; smart agriculture; intelligent agriculture; deep learning.