In today's competing business environment, the development and sustainability of an organization are meticulously tangled to the development and advancement of its employees. Career encroachment is not only a strategic persuader for employees but also an acute factor in improving organizational performance. Recognizing employees with high prospective and strategically investing in their development can build ways for improvement in productivity, innovation, and long-term success. In this paper, the prediction of employee promotion is handled by analyzing their performance based on various facts such as the number of trainings they attended, their KPI (Key Point Indicators) achievements, number of years they have served as the workforce and they training scores via our Promotion Predictive Analysis (PPA). These factors play a crucial role in determining the growth of an employee’s career and their worth as an asset to the organization. By understanding the fact that the growth of several individual employees directly comes together to influence the growth of a company, it is possible to see a direction for improvement. For this, it is imperative that the organization is able to use various techniques and tools to classify these individuals into different groups. This is done by examining the aforementioned factors and how they relate to each other, thus providing a way to strategize employee performance and career development by using a machine learning based model called SMOTEd-Boost which combined XGBoost with Synthetic Minority Oversampling technique (SMOTE) to deal with imbalanced data and classification of KPI indicators. Also our method is compared with the existing Naive Bayes and Support Vector Machines (SVM) algorithms for performance validation. Therefore, our PPA model helps others to understand the significance of the relation between an employee’s growth and an organization’s future and the need for a classification system to realize which employee’s serve the potential to be a crucial asset to the organization’s future and give them a suitable platform and also equip them with the abilities to maximize their capabilities. This would also prove to strengthen the bond between a company and their employees.

In the framework of sales forecasting, the project investigates the field of time series forecasting, with a particular emphasis on analyzing and comparing the forecasting precision of ARIMA, SARIMA, and LSTM models. Mean Squared Error (MSE), Root Mean Squared Error (RMSE), and Mean Absolute Error (MAE) are some of the performance metrics that are utilized in this procedure. It's a thorough workflow that includes data preparation, rigorous model training, and diligent assessment. To facilitate a comparative comparison of the forecasting capabilities of the models, the visualization component uses a variety of graphical representations, such as incisive line graphs and instructive bar charts. Determining which model is the most capable of providing accurate sales forecasts is the key objective. This has significant implications for improving decision-making processes in the areas of sales and inventory management. This research program helps companies to design flexible and proactive strategies that are customized to changing market landscapes. As a result, enterprises can enhance their performance and strategic resilience. This is accomplished by providing businesses with data-driven insights and rigorous forecasting tools.

In agricultural sectors like wheat, paddy, flowers, crops and maize production, insect infestations lead to significant financial losses. Swift identification of the insect species is vital to mitigate economic impacts. However, farmers often struggle to distinguish between various agricultural pests due to their limited experience and knowledge. Tackling this challenge involves harnessing advanced technologies such as the VGG-16 convolutional neural network (CNN) architecture, alongside other computational methods. VGG-16, renowned for its effectiveness in image classification tasks, can be employed to automatically extract features from the data. In our study, we propose integrating VGG-16 as part of a fusion BiLSTM network, which incorporates bidirectional long short-term memory (BiLSTM) layers, considering temporal information. This approach offers a promising solution to insect species identification in agricultural settings.

Blockchain networks serve as a transparent and secure ledger storage solution, yet they remain vulnerable to attacks. There must be some mechanism to protect the blockchain network from attacks. Among various attacks, the Distributed Denial of Service (DDoS) attack is considered severe, which is challenging to detect accurately and reliably. Machine learning techniques are used to detect the attack, which requires exploring all global attack data in a single system, which is difficult in practice. This article proposes a distributed machine learning mechanism called Federated Machine Learning for detecting the presence of DDoS attacks. But in federated machine learning the model itself can be poisoned by the malicious collaborating node which is another problem that this article solves by storing the model in blockchain and by introducing a new reputation-based miner selection procedure. The proposed framework integrates the federation of machine learning within the blockchain network framework for detecting DDoS attacks. Under the integrated framework, miners are used to train the blocks and they also participate in the machine learning training. A dynamic reputation-based miner selection mechanism that can balance exploration and exploitation is proposed for optimal miner selection, which can ensure the high accuracy of the machine learning model and improve the security of blockchain from attacks like DDoS attacks and 51% attacks. The proposed framework is tested with Random Forest, Multilayer Perceptron, and Logistic Regression machine learning algorithms. The proposed mechanism achieved maximum accuracy of 99.1% using random forest model which is superior to the existing mechanism of detection of DDoS attacks.

Cancerous cells can develop in the lungs, a condition known as lung carcinoma. One of the most deadly and challenging malignancies to cure, lung cancer often spreads to other parts of the body at an early stage. The World Health Organisation reports that for several decades, lung cancer has topped all cancers globally in terms of incidence. Early detection and treatment of lung cancer are of the utmost importance. More effective treatments, less intrusive surgeries, and longer survival times may result from this. This paper presents a method for efficiently segmenting and classifying lung carcinomas using convolutional neural networks (CNNs) for the purpose of identifying malignant tumors. Here, additional effective methods such as an adaptive median filter for preprocessing, histogram equalization, edge-based segmentation, and Convolutional Neural Network (CNN) for successful classification are used to identify the lung carcinoma. To make segmentation more efficient, we use the Adaptive Median Filter in conjunction with the Histogram Equalization technique. It is also possible to transfer the important properties to the classification algorithm by deriving them from feature similarity. When compared to another prediction model that is already in use, the suggested model has the potential to attain superior accuracy. Reduced time and effort spent on detecting lung cancer are two benefits of the enhanced performance.

In the era of social media, where information can be shared with a single click, it is important to segregate data into real and fake. Since the majority of the population uses the internet it is easy to spread misinformation. Digital information in the form of photos is one of the primary modes of communication. Unfortunately it has become increasingly easy to manipulate them. In our paper, we survey the recent proposed methods of CMFD(Copy-move forgery detection), and explained the whole CMFD process done. Particularly, we characterize a CMFD workflow for feature extraction and the matching process using key points and descriptors.

Cyber-physical systems (CPS) have revolutionized many industries by integrating physical processes, computer resources, and communication capabilities. However, the increasing connectivity and complexity of these systems have made them vulnerable to cyberattacks, including spoof attacks that distort data or inject false information into the system. Such attacks can lead to performance issues or complete system failure if not detected and addressed promptly. To address this challenge, machine learning techniques such as deep learning algorithms are being applied to simplify data processing and uncover hidden patterns that may indicate a cyberattack. In this context, a deep neural network’s structure is suggested for the detection phase, which must alert the system to the attack’s presence in its early stages.Furthermore, robust control techniques have been studied to isolate the offending agent in the leader-follower system. The control system utilizes a reputation algorithm to isolate the misbehaving agent following the attack detection phase using a deep neural network.Experimental studies have shown that deep learning algorithms can detect threats more effectively than traditional approaches, making cybersecurity easier, more proactive, less expensive, and significantly more successful. Overall, it is crucial to ensure the security of CPS to ensure their continued success and development in various industries.

Cyber-attacks are getting more sophisticated and nuanced. Intrusion Detection Systems (IDSs) are commonly used in a variety of networks to assist in the timely detection of intrusions. In recent years, blockchain technology has got a lot of attention as a way to share data without the need for a trusted third party. In particular, data recorded in a single block cannot be modified without impacting all subsequent blocks. For an effective update, an attacker will need to monitor the majority of network nodes, which is not feasible given the current network size. This work aims to create a deep learning-based IDS model with the potential of integrating blockchain technology with intrusion detection, inspired by the ability to apply blockchain in all fields. The proposed model outperforms the conventional systems with respect to accuracy in detecting the security attacks. (c) 2021 Elsevier B.V. All rights reserved.

The process of making fake images has become simple with the advancement of powerful editing softwares like Adobe Photoshop, GIMP, some of which are available for free. Image forgery detection is administered with the intention of investigating whether the image under question represents the unmodified captured image, or whether it has been forged. The copy-move forgery detection means finding the part of the picture that was forged i.e., copy-moved to another part of the picture itself. This paper presents a new image forgery detection technique where detecting the cloned digital image forgery is its focus. The proposed approach is used feature extraction methods in both detection and localization stages.

This is the era of the internet and everything has now turned online, from ordering food to ordering daily services people now cannot live without the web. Due to the pandemic the recruiting companies were also forced to conduct all of their business online. This paper discusses about the importance of web portfolios, and a one site stop for building your own portfolio from scratch along with numerous artificial intelligence features. It also discusses about chatbots and its functions. How a chatbot is made using RASA framework is also explained here.

Since technology started evolving people started using all the latest technology to make them updated and use the technology for their day-to-day life usage. In the exponential growth of technology, cloud computing plays an important role all over to world to store user’s data in it. Cloud storage provides higher storage space for less amount and can access it from anywhere, whenever you want to access the resources. In this paper, they have used a three-layer storage architecture to protect the data from unauthorized users and make the data available to the user’s all the time without any modification in it. The three-layer architecture includes the user local machine, fog layer, and the cloud layer. We propose a system where it provides confidentiality, integrity, availability of the user’s data and we are implementing new techniques like XOR-Combination to fragment and defragment the data and for creating regenerating blocks to retrieve the data successfully, hashing is used to check the integrity of the data while downloading it from the cloud server, block management is used to store the fragmented blocks in the different cloud server to make it secure and keep away it from attackers.

Blockchain which emerged in the last decade has been considered as a very powerful and potential technology since it has been applied and gained popularity from various domains like supply chain, healthcare, energy, music, food, finance, insurance, government, etc. There are several use cases in the health care where we can apply blockchain for maintaining electronic medical records, remote patient monitoring records, drug supply chain, insurance claims, hospital information system, etc. Any intentionally modified data or data misrepresentation or deception in the insurance could intend to result in unauthorized benefits. These fraudulent claims keep increasing annually and because of this, the health insurance system is on a greater disadvantage. So by using blockchain we can overcome the various challenges and issues in the current system. Our focus is to build a secure system to securely manage the health care data and also to track the insurance activities so that we can prevent health insurance fraud. We design a blockchain based solution for health care data collection andinsurance claims by deploying blockchain since it is secure, tamper proof, immutable. Thus it also gives us added security since it is impossible for anyone to modify/claim the insurance falsely since only the authorized people will beable to do so. We have used ganache blockchain to experiment the same and used solidity language for the coding.

In case of medical emergencies health insurance is extremely important in providing us with coverage for medical expenses. However, preventing data breaches and fraudulent activities involving health care data is the most difficult challenge in this sector. Transparency, which is a major issue in the health sector, is provided by Blockchain Technology. This paper uses blockchain to design a claim model that allows the system to maintain transparency between insurance companies. This proposed model eliminates the need for agents and enables direct contact between insurance companies, hospitals, and companies. In addition, it enables secure and efficient access to medical data by patients, providers, and other third parties while protecting patient privacy.

Conventional online instructive frameworks still have weaknesses when contrasted with a genuine study hall education, for example, absence of logical and versatile help, and absence of adaptable help of the introduction and input, absence of the agreeable help among understudies and frameworks. Likewise, they depend on the live information and anticipate the out comings dependent on that. This does exclude information of understudies in a foundation concentrating for some earlier years. This poses a problem for any learning and predictive algorithms to work on them. This work intends to assist the students in articulating their subject, club, project, internship, job preferences. In addition to student profiling, the venture additionally gives counsel to understudies with respect to how the profiles might be utilized to improve their scholastic and quantitative aptitude. In this regard, it is trusted that the profiles will give a valuable device to enable understudies to build up their employability. The profiling framework monitors the learning exercises and connection history of every individual understudy into the understudy profiling database model. In light of this model and along these lines the work demonstrates dynamic learning plans for individual understudies.Data analytic tools, classification techniques, and algorithms will be used to predict the outcomes of the student subject choices. Data (marks and interests) of the students will be classified into clusters upon which self-learning, predictive algorithms will be implemented to cater to students interests and needs. Regression techniques like map reduce will be used to segregate and classify data into definite datasets. It works on these four dimensions like input, comprehending, preparing and understanding. This paper gives the best way to use collaborative filtering strategies for understudy execution forecast. These strategies are frequently utilized in recommender frameworks like Netflix. The essential thought of such frameworks is to use the similitude of users dependent on their evaluations of the things in the system. We have chosen to utilize these procedures in the instructive condition to foresee understudy execution. We compute the comparability of understudies using their examination results, shown by the evaluations of their recently passed subjects.

Most of the modern day self-driving cars lack the ability to make a quick judgement based on the objects that come in front of them. We have proposed a hierarchy method in this paper by which a vehicle can decide as to how it should respond in case of a deadlock situation, i.e. when the car has no option other than crashing. We want to achieve a model of a self-driving car that can learn to keep itself on the road and also avoid obstacles in front of it and can give a priority to the different object as to which object is more valuable.

Entertainment industry is growing at a very rapid pace where a huge amount of money is being put into the making of the films. The success or failure of a movie is determined by the box office collection. The box office collection is dependent on various factors like director, actors, actresses, technicians, production house, musicians, marketing, etc. But yet it is highly influenced by the reviews and feedbacks given by the people, critics, media, etc. Due to the latest trends in the marketing field, digital media is used as a form of recommendation system where users read the reviews about a movie, or the ratings rated before making any decision to watch a movie. The existing problem that are identified with these reviews is that nowadays bots are being deployed to increase the reviews, and also fake reviews, negative reviews are also a major concern since most of the reviews are given by people without watching the movies. These problems create huge financial loss to the company, the people associated with it and also the movie industry as a whole. So in order to overcome the existing problems we are considering using blockchain technology as the future of entertainment industry. Since blockchain is a distributed network we can use it to store the reviews given by the user which is nontamperable and this in turn can help rate a movie correctly. Because of the genuine reviews stored in blockchain, movie goers can choose the correct film and make a good film successful thereby increasing the box office collection of the movie.

Blockchain is a simple and powerful platform for creating and deploying smart contracts .Its main aim is to remove the barriers in the deployment of blockchain especially for the financial sector. It provides the privacy and control required by them. It is called as smart contract because these are

Electronic voting is the use of electronic means to cast and count votes. The system must be able to offer fairness, trust, transparency, privacy, integrity, and availability of votes. The application of block chain in electronic voting seems to match the above qualities of an electronic voting system. However, recent researches have shown that block chain technology lacks anonymity. Transactions are pseudonymous and not anonymous. For this cause, a proposed method to use linkable ring signatures in block chain electronic voting system provides anonymity. With this system, a voter can verify votes with the highest level of anonymity from other voters. Voters can also tally the votes without the use of centralized third party. The proposed system will be implemented using Ethereum protocol. It uses elliptic curve cryptography to provide authentication and non-repudiation, immutability of vote records through hashing algorithms, transparency, and accessibility due to its decentralized nature as well as voter anonymity.

In India we live in a society where men and women have equal rights but as far as it goes these don’t hold up in real life scenarios and practical situations for women. Despite having all these laws and regulations to protect the rights of a woman, they still face problems. Women have grown powerful and are conquering our world’s greatest positions/frontiers but they face issues with our society’s narrow-minded behavior. Every third woman faces physical abuse/harassment in this world. Women Safety is a matter of concern and a smart phone can be used efficiently for personal security and other protection purposes. The existing apps need the user to interact with the interface and follow a sequence of steps to make them work. We are going to solve it by eliminating the user interaction with the app by providing an instant solution.