An intelligent system has recently been developed and is not a passive system for the right service. The latest system will respond to the user's request and recommend him. It is used for personalized recommendation, and the most common ways are content-based and collaborative filtering (CF). To address the issue of sparsity and collaborative filtering, we proposed a prediction technique based on a recommendation system (RS) and a collaborative filtering or rule mining (RM) strategy. To validate the prediction approach, we used consumer rating data on movies. To estimate the user's rating, utilize this Movie Lens 1M dataset. The accuracy of the method's Root Mean Square Error (RMSE) and Mean Square Error (MSE) is determined by comparing it to the real value of each movie. The present system is compared in addition to the projected RMSE and MSE.. In addition to the expected RMSE & MSE, the existing system is compared. This work also gives a comparative study of a variety of similarity measures, including the Cosine, Cityblock, Chevyshev, Braycurtis, and Spearman correlation. Rule mining with the Eclat rule is completed for the ideal recommendation on a rating matrix. Recall, Precision, & F1 score measure is used to assess recommendation accuracy. From the experimental results, we can show that Cityblock and Chevyshev measure based Eclat rule mining achieved better precision and F1 Score in comparison to other measures.

Internet and Multimedia technologies have become our daily needs. Hence it has become a common practice to create copy, transmit and distribute digital data. Obviously, it leads to unauthorized replication problem. Digital image watermarking provides copyright protection to image by hiding appropriate information in original image to declare rightful ownership. The study focuses on overview of several Transform Domain watermarking methods with detail mathematical formulae, their implementations, strengths and weaknesses. Different color models in image processing with their comparative study are discussed in this paper. The generalized algorithms are presented for DWT, DCT-DWT, SVD, DWT-SVD approaches. Comparative study of various researchers’ work on color image security techniques used for watermarking is also presented in this paper.

Electrical power networks are built and maintained to fulfill the fluctuating demand for electricity. The running expenses of a power grid must be kept as low as possible. Dynamic Economic Load Dispatch (DELD) is a strategy for scheduling power generator outputs based on load requirements and managing the power system as efficiently as feasible. To put it another way, the fundamental goal of economic load dispatch is to distribute the maximum amount of power provided by varied units at the lowest possible cost while fulfilling all system constraints. Under regular operating conditions, power plants in a realistic power grid are not positioned at the same distance from the centre of loads, and their fuel prices fluctuate. Furthermore, the whole load demand is exceeded by the producing output, resulting in losses. However, the most important considerations for a system should be security, cost effectiveness, and dependability. This paper proposes a modified particle swarm optimization (MPSO) technique for dealing with equality and inequality constraints in ELD problems using Gaussian and Cauchy probability distributions. The MPSO method introduces a new diversification and intensification strategy into the particles, preventing the PSO algorithm from achieving premature convergence.

Currently, the Brain-Computer Interface (BCI) is an interesting research area whose goal is to create an interaction channel between the system and the person's brain. It provides a direct means of turning brainwaves into physical impacts that does not require the use of muscles. BCIs are customized devices that allow users to control computer programs with their brain waves. With the introduction of consumer-grade electroencephalography (EEG) equipment, brain-controlled systems began to find uses outside of the medical realm, opening up a plethora of research options in the field of Human-Computer Interface (HCI). In this paper discuss the application of BCI for finding more accurate results for the nerve diseases Amyotrophic Lateral Sclerosis (ALS). Here in this paper implement machine learning approach to finding the results with help of teaching learning based optimization techniques.

Medical science has grown tremendously in the last decade. However, our country's healthcare system is unable to keep up with the demand of an increasing population. An increasing number of people prefer private hospitals over government hospitals given the availability of physicians, equipment, diagnostics, and services, among other things. Today's health-care industry creates and produces vast or massive amounts of datasets, necessitating the development of specialized tools for the effective and accurate extraction of knowledge or information from stored data. Data mining provides a variety of methodologies and tools for data diagnosis in the health care business. Some data is based on clustering, while others are based on categorization, association rule analysis, regression, and so on. Data classification is one of the most extensively utilized data mining approaches in healthcare organizations. It is an example of supervised classification methods. Classification algorithms are also used to anticipate the therapy cost of healthcare services since data is separated into calls and recognized according to their provided aim or output. Current data mining and machine learning approaches are particularly effective for diagnosing patients and storing massive amounts of data. Some optimization strategies are also employed to increase the detection rate and diagnosis findings of illnesses. Modern medicine is a one-of-a-kind combination of a patient, healthcare experts, and technology. In this dissertation, we proposed a new model based on classification methods including support vector machines, neural networks, and optimization methods, which is a bi-logically inspired method for improving classifier results in terms of some performance characteristics such as accuracy, precision, recall, and so on. We measure and improve for all performance parameters for various datasets such as heart patients, liver patients, and cancer patients. The dataset for all patients is collected from the UCI machine learning repository, which provides the legitimate dataset for the study activity, and the simulation program is MATLAB. Our experimental results in this dissertation demonstrate that our approach has a higher detection rate of categorization for performance characteristics than other current techniques.

The expected rapid development sought after higher information rates and further developed administrations in cell networks infer that the cell network design should be reconsidered. It implies that there is need for new strategies to expand the range use. Method dependent on broadening existent arrangements with CR usefulness joined with specially appointed organization strategies is needed to help the expected interest for higher throughputs. The main viewpoints are the need to grow precise and effective administration systems to administer the unique range access of CR. The CR territory has zeroed in among others on new answers for energy and phantom effectiveness in remote correspondence (green activity), range task for quick range access, directing and handover instruments, dynamic and expectation calculations to limit the effect of optional clients (SUs) on essential clients (PUs). In this paper discussed the different spectrum sensing techniques in cognitive radio network.

Data mining is a technique widely employed in the medical industry to retrieve disease-related data. Heart disease is the most common human ailment. The big data created during cardiac disease prediction is too complicated and large to analyze and evaluate using existing approaches. Using data mining methodology reduces the time required to accurately anticipate illness. To diagnose and extract meaningful information from the large dataset, this thesis presents a hybrid of SVM and KNN. To solve the problem of cardiac disease, the simulation program MATLAB is employed. Coronary infection is a vicious infection that affects millions of individuals worldwide. Early detection is vital given the high loss of life charges and large number of people who suffer from cardiovascular sickness. Studying a disease isn't always easy. Using a recovery investigation form to an ai for coronary disease gives more security than mining human administration data. DB mining today, clinical test findings are frequently decided based on specialists' insight and expertise rather than the rich data available in several vast databases. This method frequently causes unintentional predispositions, lapses, and high medical costs, affecting the quality of patient care. Many doctors' offices now use secure data management systems to store patient and social security information. These data frameworks often produce a lot of data in different formats, such as statistics, for clinical decision making. The heart is an essential organ. This extensive database is occasionally used by the heart to operate efficiently. If the heart is not working properly, it will affect the brain, kidneys, etc. It is a pump that circulates blood throughout the body. Inefficient blood circulation harms organs like the brain, and when the heart stops pumping, death happens within minutes. Heart efficiency is vital to life. Heart disease is a disease of the heart and its blood vessels. Several variables raise the risk of heart disease. These methods and strategies are used to find knowledge from databases. Some data mining methods describe.