Search

The superiority of Convolutional Neural Networks (CNNs) largely relies on their architectures that are usually manually crafted with extensive human expertise. Unfortunately, such kind of domain knowledge is not necessarily owned by every interested user. Data mining on existing CNNs can discover useful patterns and fundamental comments from their architectures, providing researchers with strong prior knowledge to design effective CNN architectures when they have no expertise in CNNs. There are various state-of-the-art data mining algorithms at hand, while there is only rare work on mining CNN architectures. One of the main reasons is the gap between CNN architectures and data mining algorithms. Specifically, the current CNN architecture descriptions cannot be exactly vectorized to feed to an data mining algorithm. In this paper, we propose a unified approach, named ArcText, to describing CNN architectures based on text. Particularly, four different units and an ordering method have been elaborately designed in ArcText, to uniquely describe the same CNN architecture with sufficient information. Also, the resulted description can be exactly converted back to the corresponding CNN architecture. ArcText bridges the gap between CNN architectures and data mining researchers, and has the potential to be utilized to wider scenarios.

The Gunung Maligas District Office is a government agency tasked with running a government program, namely the Social Assistance Receipt program, to run the social assistance program, many residents complain that they do not receive assistance, while some residents who are considered capable actually get assistance, where each aid program is have different criteria in determining the recipient. Due to the large number of existing aid programs with different criteria in determining the acceptance of the aid program, of course, local government staff will have difficulty in conducting the selection process. So we need a system that is able to help local government staff to more easily determine the recipients of the social assistance. Based on the historical data of beneficiaries, recommendations for the classification of beneficiaries can be made that will assist government staff. Classification can be done using the C4.5 algorithm. In this study, it has parameters, namely, occupation, income, housing conditions and number of dependents. By applying the C4.5 data mining algorithm, it is hoped that it will make it easier and faster for government staff to determine the recipients of social assistance at the Gunung Maligas District Office.

High utility itemset (HUI) mining is a powerful data mining technique to discover profitable patterns. The utility of an item is computed by using two measures named quantity and per-unit profit. All existing HUI mining algorithms consider a single value of external utility (per unit profit) for the entire database. However, the per-unit profit of items might fluctuate over time in many applications. This research introduces three novel strategies to comprise the external utilities of items as input for the HUI mining algorithm. Traditional HUI mining algorithms have been developed for the standalone system and do not fit for big data processing due to the limited computing resources (CPU, memory). Big data are efficiently processed on distributed frameworks like Apache Hadoop, Spark, etc. This paper introduces a distributed HUI mining algorithm named Spark-based Top-k high utility itemset (k-SHUI) miner. We also propose a fair load distribution strategy to divide the search space equally among the cluster nodes. The k-SHUI produces top-k HUIs without the requirement of the minimum utility threshold. We conducted extensive experiments on six real-life datasets to compare the proposed algorithm's performance with the existing algorithm. The experimental results demonstrate that the proposed algorithm outperforms the existing algorithms.

Background: The computerised classification and prediction of heart disease can be useful for medical personnel for the purpose of fast diagnosis with accurate results. This study presents an efficient classification method for predicting heart disease using a data-mining algorithm. Methods: The algorithm utilises the weighted support vector machine method for efficient classification of heart disease based on a binary response that indicates the presence or absence of heart disease as the result of an angiographic test. The optimal values of the support vector machine and the Radial Basis Function kernel parameters for the heart disease classification were determined via a 10-fold cross-validation method. The heart disease data was partitioned into training and testing sets using different percentages of the splitting ratio. Each of the training sets was used in training the classification method while the predictive power of the method was evaluated on each of the test sets using the Monte-Carlo cross-validation resampling technique. The effect of different percentages of the splitting ratio on the method was also observed. Results: The misclassification error rate was used to compare the performance of the method with three selected machine learning methods and was observed that the proposed method performs best over others in all cases considered. Conclusion: Finally, the results illustrate that the classification algorithm presented can effectively predict the heart disease status of an individual based on the results of an angiographic test.

With the advent of the big data era, information presentation has exploded. For example, rich methods such as audio and video have integrated more information, but with it, a lot of bad information has been brought. In view of this situation, this paper relies on data mining algorithms, builds a multimedia filtering system model for massive information, and integrates content recognition, packet filtering, and other technologies to match the two to ensure the integrity and real time of filtering. Practice results prove that the method is effective.

Next-term grade prediction is a challenging problem. The objective of this problem is to predict students grades in new courses, given their grades in courses they have previously taken. Adopting various machine learning algorithms is a very common and straightforward approach to tackling this problem. However, such models are very difficult to interpret. That is, it is difficult to explain to a student (or a teacher) why the model predicted grade B for a given student for example. In this work, we shed light on the importance of building interpretable models for educational data mining tasks. Specifically, we propose a novel interpretable framework for multi-class grade prediction that is based on an optimal rule-list mining algorithm. Additionally, we evaluate our proposed framework on two private datasets and compare our results with baseline models. Our findings show that our proposed framework is capable of achieving higher prediction and interpretability values when compared to black-box models.

Data mining techniques are included with Ensemble learning and deep learning for the classification. The methods used for classification are, Single C5.0 Tree (C5.0), Classification and Regression Tree (CART), kernel-based Support Vector Machine (SVM) with linear kernel, ensemble (CART, SVM, C5.0), Neural Network-based Fit single-hidden-layer neural network (NN), Neural Networks with Principal Component Analysis (PCA-NN), deep learning-based H2OBinomialModel-Deeplearning (HBM-DNN) and Enhanced H2OBinomialModel-Deeplearning (EHBM-DNN). In this study, experiments were conducted on pre-processed datasets using R programming and 10-fold cross-validation technique. The findings show that the ensemble model (CART, SVM and C5.0) and EHBM-DNN are more accurate for classification, compared with other methods.

Renewable energy sources are gaining a significant research attention due to their economical and sustainable characteristics. In particular, solar power stations are considered as one of the renewable energy systems that may be used in different locations since it requires a lower installation cost and maintenance than conventional systems, despite the fact that they require less area. In most of the small generating stations, space occupancy is controlled by placing the equipment on an open terrace. However, for large-scale power generating stations, acres of land are required for installation. Human employers face a challenging task in maintaining such a large area of power station. Through IoT and data mining techniques, the proposed algorithm would aid human employers in detecting the regularity of power generation and failure or defective regions in solar power systems. This allows performing a quick action for the fault rectification process, resulting in increased generating station efficiency.

Data preprocessing is an essential task to prepare suitable target datasets to apply statistical and data mining algorithms. It has become one of the complex segment in Web usage mining due to the ...

At present, PBL English (foreign language) micro audio and video teaching mode is gradually rising, and the application of intelligent algorithm in it has gradually attracted people’s attention. This paper studies PBL English micro audio and video teaching mode based on human-computer interaction and micro audio and video recognition and data mining algorithm. Through the extraction and selection of acoustic features of speech emotion, the relationship between emotion and each feature parameter is explored, and the acoustic feature with the strongest emotion discrimination ability is selected. Using the machine learning method of support vector machine, a speech emotion recognition model with high recognition rate is constructed and applied to the mobile speech emotion recognition system, and the algorithm is verified by experiments. The experimental results show that the prediction accuracy is 70%. The algorithm proposed in this paper has good recognition effect, strong human-computer interaction ability and good effect. The research of this paper will contribute to the development of English (foreign language) micro audio and video teaching mode, optimize the human-computer interaction mode of English (foreign language) micro audio and video teaching, and improve the teaching autonomy.

Beer is considered one of the top three most popular drinks, being consumed all over the world. During the last few decades, discrimination of beverages and food products has gained attention with many application research studies based on chemical parameters and chemometric or machine learning algorithms. However, no reviews about the evaluation of beers have been reported. Therefore, this review presents applications of beer classification among brands, styles and types, aging, origin, and the prediction of quality attributes of interest based on chemometric, machine learning methods, and chemical parameters. After analyzing the literature, it was found that chemometric and machine learning methods are successful tools for qualitative and quantitative examination of beers. However, more work needs to be done to evaluate machine learning methods and data mining algorithms, such as sampling, feature selection, and advanced classification algorithms.

This article considered only natural types of texture and then applying the Gabor filter for better classifications. The concept used is to discard the stochastic features to avoid any mixing of feature vector while it is extracting from the image dataset. The proposed approach has considered the Gabor filter for texture recognition primarily but with the combined method of spatial width and orientation to get the optimal alignment, this optical alignment mine the maximum feature vector by applying the REP algorithm over the data mined from the texture. This will result in better accuracy in the results. Initially, the frequency response over the surface due to applying Gabor filter has been calculated and then the work proceeded in a manner that first natural images are loaded into the MATLAB tool then it is preprocessed, and then final classifications are performed for final results. The primarily concentrated over texture information of image datasets rather than the multispectral information along with REP regression algorithm to do actual mining of feature vectors. Unlike the conventional approach of the Gabor filter, this article focuses on the variance and spatial relationship between two or more than two pixels. The deviation calculated is used for normalizing the feature vectors, and the accuracy can be hence increase using the proposed commuted technique.

In order to address the problems of low precision, long time-consuming and low recall rate in mining complex attribute medical data in medical information, an intelligent mining algorithm for complex attribute medical data based on deep learning is proposed. Discretized medical data with complex attributes and converted it into a data type suitable for deep learning research, the convolutional neural network is used to analyze the association mapping relationship between complex attribute medical data sets and extract association rules of data. According to the degree of association between complex attribute medical data sets in multi-dimensional subspace to realize the effective mining of complex attribute medical data. The results show that the proposed algorithm takes less time and can extract association rules accurately, the data priority control efficiency is higher, the data mining accuracy is better, and the data mining recall rate is much higher than other methods, which verifies the feasibility of the proposed algorithm.

Data stream is the huge amount of data generated in various fields, including financial processes, social media activities, Internet of Things applications, and many others. Such data cannot be processed through traditional data mining algorithms due to several constraints, including limited memory, data speed, and dynamic environment. Concept Drift is known as the main constraint of data stream mining, mainly in the classification task. It refers to the change in the data stream underlining distribution over time. Thus, it results in accuracy deterioration of classification models and wrong predictions. Spam emails, consumer behavior changes, and adversary activates, are examples of Concept Drift. In this paper, a Concept Drift detection model is introduced, Concept Drift Detection Model (CDDM). It monitors the accuracy of the classification model over a sliding window, assuming the decline in accuracy indicates a drift occurrence. A modification over CDDM is a weighted version of the CDDM as W-CDDM.Both models have evaluated against two real datasets and four artificial datasets. The experimental results of abrupt drift show that CDDM, W-CDDM outperforms the other models in the dataset of 100K and 1M instances, respectively. Regarding gradual drift, the W-CDDM overtook the rest in terms of accuracy, run time, and detection delays in the dataset of 100 K instances. While in the dataset of 1M instances, CDDM has got the highest accuracy using the NB classifier. Moreover, W-CDDM achieves the highest accuracy on real datasets.

High Utility Itemset Mining aims at extracting high utility patterns from a given database. Most of the mining algorithms are designed for static datasets. However, the real-life datasets continuously evolve with incremental data. In this paper, we present a mining algorithm which uses efficient data structures and pruning conditions to mine and maintain high utility patterns across incremental datasets. Several experiments are conducted to show the efficacy of the proposed algorithm.

This paper proposes an advanced personality classification and professional profile prediction system. The system grasps from the personality patterns by using advanced data mining to mine user characteristics data. This E based on its past classifications from the data mining algorithms, and it classifies or predicts the user's personality. Several user characteristics and behaviors are analyzed by this system and based on the patterns observed, it stores its own user characteristics patterns which then are stored in a database. On the basis of personality data which was stored by classification of previous user data, the system predicts user personality. This can further be used for job prediction, subject selection, etc. This system is useful for hiring purposes of companies as well as for characterization of team members already working in a professional environment.

Heart Disease refers to a broad range of heart-related health problems. Heart disease is currently the world's most serious public health issue. Many organizations have made extensive use of data mining. Data mining in healthcare is becoming trendy, if not extremely important. The health sector nowadays produces a significant volume of complex data about individuals, diagnosis of diseases, clinical notes, medical equipment, and so on. The objective is to know about the various data mining methods that have evolved to forecast heart problems. According to the findings, a Random forest with 15 features outstripped all such data-mining methods.

This paper proposes a framework for computer-aided talent evaluation. Intelligent evaluation is added on the basic questionnaire evaluation, in which questionnaire evaluation is used as explicit evaluation, and the results are quantitatively scored based on industry development questionnaires, scales and explanations. Intelligent evaluation is mainly based on data mining algorithm. As an implicit evaluation, it aims at processing talents’ characteristic engineering, forming users profile and recommending jobs. The final evaluation report is obtained by integrating different evaluation methods. This paper explains the composition of the evaluation framework in detail, and studies its application cases in evaluating talents’ digitalization capability.

Purpose This paper aims to propose a structured model based on a data mining algorithm that can calculate, based on business association (BA) attributes, the probability of micro and small enterprises (MSEs) becoming a new member of a BA. Another goal is the probability of a BA attracting new members. Design/methodology/approach As a methodological procedure, the authors used the Naive Bayes data mining algorithm. The collected data were analyzed both quantitatively and qualitatively and then used to define the model, which was tested randomly, while allowing for the possibility of future validation. Findings The findings suggest a structured model based on a data mining algorithm. The model can certainly be used as a management tool for BAs concentrating their efforts on those businesses that are certainly potential new recruits. Further, for an MSE, it serves as a means of evaluating a BA, indicating the possible advantages in becoming a member of a particular association. Research limitations/implications This paper is not intended to be generalized, considering that it only analyzes the BAs of Rio Grande do Sul, Brazil. In this way, when applying this model to other situations, the attributes listed here can be revised and even modified to adapt to the situation in focus. Practical implications The use of the proposed model will make it possible to optimize the time of BA managers. It also gives MSE greater reliability in choosing BA. Social implications Using this model will provide better decision-making and better targeting, thus benefiting both the BAs and the MSEs, which can improve their management and keep jobs. Originality/value This paper contributes to the literature because it is the first to connect BAs, MSEs and Naive Bayes. Also, this study helps in better management for BA managers in their daily activities and provides a better choice of BA for MSE managers. Also, this study contextualizes BAs, MSEs and data mining in an objective way.

One of the most common health problems that correlated to serious complications is chronic kidney disease. Early detection and treatment can save it from progression. Machine learning is one tool that used historical data to improve future decision about prediction of chronic kidney disease. The aim of this work is to compare the performance of six different models based on accuracy, sensitivity, precision, recall. In this study, the experiments were conducted on 158 records downloaded from UCI repository. Six algorithms ( K-Nearest Neighbor, Naïve Bayes, Support Vector machine, Logistic Regression, Decision Tree, and Random Forest ) were implemented on data after preprocessing stage. Evaluation of models resulted in Naïve Bayes and Random Forest accuracy 100%, Sensitivity 100%, Specificity 100%, precision 100 %, Recall 100% respectively. It is concluded that Naïve Bayes and Random Forest are better than other models.

Background Data mining is a technique widely used in medicine for the medical diagnosis, treatment and prognosis. Amongst the medical applications of data mining techniques, the most important are the construction of models comprising the key factors in disease diagnosis, evaluation of disease-related data patterns, monitoring of patient health status, and forecasting patients’ future status. The purpose of this study was to assess key factors in the onset/progression of asthma central to accurate diagnosis of the disease. Methods The survey was designed as a cross-sectional descriptive study in the summer of 2017. We selected patients diagnosed with asthma (n = 300) and individuals without asthma (n = 80) referred to clinics and hospitals in Tehran (Iran), Mofid Hospital (Tehran) and the Children's Medical Center (Tehran). A questionnaire was used to collect complementary medical information from the selected individuals. The recorded information was then evaluated and 200 of these cases were selected for further analysis. The surveyed data was analyzed using MATLAB and RapidMiner software and the results were assessed using a decision-tree algorithm. Results and Discussion The key disease factors and their interrelation were analysed in the present study. Shortness of breath, wheezing, repeated respiratory attacks, cough, sleep disorder and facial bruising are common symptoms of asthma from the mildest to the most severe forms. Family history was also found to be an important factor in asthma risk for children. The findings of the study are consistent with those of expert physicians, with confirmation of the ability of data mining to generate knowledge from experimental data and to establish a consensual diagnosis.

Sensors and mobile phones shine in the Big Data area due to their capabilities to retrieve a huge amount of real-time data; which was not possible previously. In the specific field of healthcare, we can now collect data related to human behavior and lifestyle for better understanding. This pushed us to benefit from such technologies for early miscarriage prediction. This research study proposes to combine the use of Big Data analytics and data mining models applied to smartphones real-time generated data. A K-means data mining algorithm is used for clustering the dataset and results are transmitted to pregnant woman to make quick decisions; with the intervention of her doctor; through an android mobile application that we created. As well, she receives recommendations based on her behavior. We used real-world data to validate the system and assess its performance and effectiveness. Experiments were made using the Big Data Platform Databricks.

Today, development of technology and the use of modern medical equipment and update technology produce massive amounts of stored information in the medical database. Analysis and discovery of knowledge from medical database is difficult due to the high volume of data and is requires a newer technology that data mining technologies to achieve this important to help its powerful algorithms. Data mining techniques with extract knowledge and the unseen patterns from huge volumes of data and build models related to medical databases, designed decision support system that help decision-making to Medical. The main objective of this paper has been to examine how to apply data mining techniques to predict and diagnosis of Down syndrome disease based on the medical information of 200 patients referred to medical laboratories in the country by using data mining algorithm CHAID in Rapid Miner software. The results showed that the CHAID algorithm with 100% accuracy has ability diagnosis of Down syndrome disease.

Frequent itemset mining serves as the main method of association rule mining. With the limitations in computing space and performance, the association of frequent items in large data mining requires both extensive time and effort, particularly when the datasets become increasingly larger. In the process of associated data mining in a big data environment, the MapReduce programming model is typically used to perform task partitioning and parallel processing, which could improve the execution efficiency of the algorithm. However, to ensure that the associated rule is not destroyed during task partitioning and parallel processing, the inner-relationship data must be stored in the computer space. Because inner-relationship data are redundant, storage of these data will significantly increase the space usage in comparison with the original dataset. In this study, we find that the formation of the frequent pattern (FP) mining algorithm depends mainly on the conditional pattern bases. Based on the parallel frequent pattern (PFP) algorithm theory, the grouping model divides frequent items into several groups according to their frequencies. We propose a non-group PFP (NG-PFP) mining algorithm that cancels the grouping model and reduces the data redundancy between sub-tasks. Moreover, we present the NG-PFP algorithm for task partition and parallel processing, and its performance in the Hadoop cluster environment is analyzed and discussed. Experimental results indicate that the non-group model shows obvious improvement in terms of computational efficiency and the space utilization rate.

Introduction Widespread adoption of Electronic Health Records (EHR) increased the number of reported disease association studies, or Phenome-Wide Association Studies (PheWAS). Traditional PheWAS studies ignore visit type (i.e., department/service conducting the visit). In this study, we investigate the role of visit type on disease association results in the first Visit-Wide Association Study or ‘VisitWAS’. Results We studied this visit type effect on association results using EHR data from the University of Pennsylvania. Penn EHR data comes from 1,048 different departments and clinics. We analyzed differences between cancer and obstetrics/gynecologist (Ob/Gyn) visits. Some findings were expected (i.e., increase of neoplasm diagnoses among cancer visits), but others were surprising, including an increase in infectious disease conditions among those visiting the Ob/Gyn. Conclusion We conclude that assessing visit type is important for EHR studies because different medical centers have different visit type distributions. To increase reproducibility among EHR data mining algorithms, we recommend that researchers report visit type in studies. Electronic supplementary material The online version of this article (10.1186/s13040-019-0203-2) contains supplementary material, which is available to authorized users.

The search for information associated with qualitative data is usually done using data mining algorithms, the presented research analyzes data of patients with essential hypertension (HTA), patients who have developed hypertension but there is no clear reason why it has occurred. In this research, a search of behavioral patterns was performed in the data associated with the clinical records of 8470 patients using the Random Forest algorithm. As a case study, the proposal focuses on finding the relationship between the different pathologies or factors associated to Hypertensive patients (other diseases for example). The findings validate the right use of the algorithm due to the results obtained agrees with the knowledge defined and validated in the literature. Thus, trivial knowledge can be obtained with the algorithm used. However, non-trivial knowledge was also obtained given the analysis performed on a total of 4408 data of female patients and 4062 of male patients showed a great difference between the factors or pathologies that a patient presents when classified according to their sex, thus another deep study must be carried out closely with experts in the area of the health as future research.

Data mining is the practice of examining unknown patterns of data according to diverse viewpoints for classification into valuable information, which is composed and gathered in collective areas, such as data warehouses.For effective analysis, data mining algorithms enabling business decision making and other information necessities to eventually cut costs and raise revenue. Sentiment analysis is the method of defining the emotional tone behind a sequence of words, used to gain an accepting of the attitudes, opinions and emotions conveyed within an online mention. Sentiment analysis is tremendously useful in social media observing as it allows us to gain a synopsis of the broader public opinion behind definite topics. The applications of sentiment analysis are extensive and influential. The ability to abstract insights from social data is a practice that is being broadly adopted by organizations across the world. In this paper, we focused on sentiment analysis on the twitter data.

Wildfires, whether natural or caused by humans, are considered among the most dangerous and devastating disasters around the world. Their complexity comes from the fact that they are hard to predict, hard to extinguish and cause enormous financial losses. To address this issue, many research efforts have been conducted in order to monitor, predict and prevent wildfires using several Artificial Intelligence techniques and strategies such as Big Data, Machine Learning, and Remote Sensing. The latter offers a rich source of satellite images, from which we can retrieve a huge amount of data that can be used to monitor wildfires. The method used in this paper combines Big Data, Remote Sensing and Data Mining algorithms (Artificial Neural Network and SVM) to process data collected from satellite images over large areas and extract insights from them to predict the occurrence of wildfires and avoid such disasters. For this reason, we implemented a methodology that serves this purpose by building a dataset based on Remote Sensing data related to the state of the crops (NDVI), meteorological conditions (LST), as well as the fire indicator “Thermal Anomalies”, these data, were acquired from “MODIS” (Moderate Resolution Imaging Spectroradiometer), a key instrument aboard the Terra and Aqua satellites. This dataset is available on GitHub via this link ( https://github.com/ouladsayadyounes/Wildfires ). Experiments were made using the big data platform “Databricks”. Experimental results gave high prediction accuracy (98.32%). These results were assessed using several validation strategies (e.g., classification metrics, cross-validation, and regularization) as well as a comparison with some wildfire early warning systems.

s-Diabetes mellitus, also called diabetes, is a metabolic disease that causes increased blood sugar. In diabetic patients, insulin is taken, as this hormone transfers sugar into the cells from the blood. At an early stage, prediction of diabetes can lead to better treatment. It is possible to predict diabetes using data mining, deep learning, machine learning, etc. Data mining is widely used for prediction, prognosis, analysis, risk factors. The proposed research work is primarily focused on the prediction of diabetes in patients in this paper, based on highest accuracy. It is possible to use classification methods based on diabetic data to predict the outcome or to discover whether or not the patient is affected. Throughout this document, a predictive model has been developed by utilizing five data mining classification model, like Naive Bayes, SMO, Multiclass, Random Forest, IBK, to predict early stage diabetes and calculate accuracy. All five algorithms are measured on different metrics, i.e, Accuracy, Recall, Precision.

The International Maritime Organization (IMO) enforced stricter sulphur abatement regulations since shipping emission has become one of the most major cause of the atmospheric pollution. Experts from the industry and academicians try to find the balanced solution among low-sulphur fuel, clean energy, and purposely fit scrubber by conventional statistical methods however failed to reach a satisfying conclusion. In addition, maritime datasets are usually massive, multi-source, and heterogeneous, it seems imperative for the maritime industry to adapt to the worldwide trend of intellectualisation and promote sustainable development.This work delineates and compares three main sulphur abatement solutions for ships through a thorough investigation of the current research state, and proposes a new framework based on a fusion model using modern big data and data mining algorithms. This work identifies and summarises major factors (with high impacts) in sulphur abatement solutions in the ocean shipping industry and integrate those high-level impacting factors to the proposed fusion model. The proposed framework can be optimised and utilised in determining suitable solutions for different ships, as well as shipping routes.

In order to solve the problem of low accuracy of data mining in traditional data mining platforms, the construction of data mining platform in the cloud computing environment is proposed. Based on the data mining platform architecture design in the cloud computing environment, and the data collection and scheduling center design, completed the hardware design of the data mining platform in the cloud computing environment; Relying on the cloud strategy design of data mining and the parallel design of mining algorithm, the software design of the data mining platform in the cloud computing environment is realized and the construction of the data mining platform in the cloud computing environment is completed. Experimental data show that the proposed data mining platform, compared with the traditional mining platform, improves the mining accuracy by 47.29%, which is suitable for data mining in the cloud computing environment.

Publish/subscribe distributed systems are often used in critical applications. It is necessary to monitor their running patterns in real time to detect abnormal status. Therefore, identifying the normal running pattern is the precondition of monitoring publish/subscribe distributed systems. Based on Apriori algorithm, this paper presents a weighted frequent itemset mining algorithm for running pattern recognition of publish/subscribe distributed systems. By introducing the transaction matrix, the algorithm only needs to scan the transaction database once. By weighting the items from two aspects of influence and frequency, the support of the items with few occurrences but much importance can be improved, so that the running pattern containing small frequency events can be mined out. Experimental results show that the algorithm can effectively mine the running patterns, and has better performance than Apriori algorithm and FP-growth algorithm.

Internet marketing requires a personalized marketing strategy. In this study, the application of data mining in personalized Internet marketing was studied. Based on the mining algorithm, a personalized marketing method was designed. Through the calculation of frequent closed item sets and support counts of positive and negative samples, the interval with a high success rate for marketing was obtained. With performance analysis, it was found that the marketing method of this study successfully improved the rate about 8% compared with the traditional marketing method. This method would have better performance under the smaller interval number and smaller minimum success number. After the actual application in telecommunication enterprise A, it was found that after adopting the marketing method of this study, the marketing success rate of A enterprise increased from 2.72–6.31%, which indicated the effectiveness of the method. The research results in this study proved the role of data mining algorithms in Internet marketing, which was conducive to the further application of mining algorithms in personalized marketing.

Applying machine learning and data mining algorithms over data distributed in multiple sources is challenging. One complication is to perform data analysis without compromising personal information, which is a primary concern in healthcare applications. Another issue involves communication overhead incurred from the transfer of raw data from one party to others for conducting centralized data mining. In healthcare applications, we are particularly interested in running data mining algorithms over big data without disclosing sensitive information about data subjects due to privacy and legal concerns. In this paper, we consider the classification problem and show how the Extremely Randomized Trees (ERT) algorithm could be adapted for settings where (structured) data is distributed over multiple sources. We propose the Privacy-Preserving Distributed ERT approach for privacy-preserving utilization of the ERT algorithm in a distributed setting. To the best of our knowledge, this is the first application of the ERT algorithm in the distributed setting, with privacy consideration (without sharing the raw data or intermediate training values), without any loss in classification performance.

Nowadays, the Health care industry is one of the fastest-growing industries. As we already know, health care has researched very widely, introducing many medical data that is not easy to mine. Data mining is an approach that helps to discover essential data from massive data or collection of data. So, in medical Science, there is a need for tools that help analyses the data, extract the significant result from massive data, and discover efficient use of information. Generally, three things are mandatory in medical for every patient. First is patient details, diagnosis and medications. Converting these data into a basic pattern for predicting the patient disease helps in early diagnosis. This research mainly focuses on the data mining approach, which is widely considered in the medical field.

Eye diseases are very common diseases in the medical field. High intraocular pressure is one of serious eye diseases, which leads to the loss of vision in the affected eye. Thus, how to effectively discover the hidden risk factors from large data of high intraocular pressure is a very significant and challenging issue. Data mining algorithms (such as classification and association rules) can discover the hidden knowledge from large-scaled data sets in many real-world applications especially in the medical field. Therefore, this paper focuses on finding out the risk factors hidden in high intraocular pressure effectively and efficiently by using J48, and Apriori algorithms. In terms of these data mining algorithms, we can discover some interesting and useful rules in the data set of eye disease. Results of the study discover some risk factors of high intraocular pressure which have been proved by medical trials. The experimental study demonstrates the effective and usability of data mining techniques in Medical field. Meanwhile, this study discovers some new risk factors for high intraocular pressure which have not been discovered previously by studies based on medical trials. This knowledge will contribute to early detection and preventing blindness by recognizing risk factors for high intraocular pressure.

The health care industries have a very large amount of dataset that contains some abstract knowledge or information which is very useful for creating good decisions. To provide good Results on data, highly advanced data mining algorithms are used. This Proposed System predicts the chances of getting heart disease in advance. In this System, a heart disease prediction using machine learning is developed using KNN and decision Tree algorithm for guessing the risk level of heart disease. The Proposed System uses 13 medical parameter such as age, sex, fast blood sugar, chest pain, etc., The results of the proposed system gives the chances of getting heart disease in advance in terms of percentages and also displays the accuracy level of the above two algorithms.

Photos shared by users of online social network often associated with tags, texts, geographic and time information. These data are ideal research resource which can be used by researchers to observe people's interests and behaviors. They can also be used to analyze people's travel behaviors and interests because it can reflect people's travel activities and experiences directly. This project aimed to consolidate people into groups based on their travel preferences of travel season, travel time span, scenery type of travel location and then analyze popular travel places for different people groups. The travel preferences were analyzed from extracted data applying PHP language and Naive Bayes classification approach. According to the analyzed data, people were divided into groups. Then, DBSCAN clustering was used to get popular locations based on the different groups. As a result, 80 thousand photo data of 5 cities have been collected from Flickr by Phython. It has been showed that people's preference of season in most of the mined cities was spring and summer and the preference of time span was short. Moreover, mined people has been divided into 16 groups according to their preference and based on the groups popular locations have been calculated by DBSCAN clustering. Theses locations have been provided in a recommendation website which was built in this project. For the future work, mining Flickr data associated with other travel social network sites could be a good way to build a more effective recommendation system.

The application of data mining in teaching has entered the stage of development. During the epidemic, colleges and universities have accumulated a large amount of online teaching statistical data. These data can be used to establish a classification model for predicting student performance. In this paper, the Naive Bayes algorithm is improved to solve the problem of underflow when the data feature values are too large. The student performance prediction classification model is constructed, and the classification efficiency and accuracy are improved to a certain extent. The improved model is used to predict and warn the performance in the mid-term stage to prevent The phenomenon of a large proportion of missing subjects, thereby ensuring the quality of students’ learning throughout the semester.

The development of international agriculture trade during the COVID-19 pandemic has encountered significant challenges. The processing of international agricultural trade data using machine learning techniques needs to be improved to perform effective analysis of agricultural trade. An essential issue for international agricultural trade is the accurate yield estimation for the numerous crops involved in international trade. Data mining techniques are the necessary approach for accomplishing practical and effective solutions for this problem. This paper combined the bidirectional encoder representations from transformers (BERT) model to conduct data mining and developed a trade data analysis system with efficient data analysis capabilities. Our results indicate that our model does reasonably well and obtains adequate information in deciding international agricultural trade. It can also be instrumental for policy and decision-making regarding international agricultural trade.

This paper describes the results of experiments in predicting students’ performance on a massive open online course (MOOC). Grade reports from MOOC “Data management” on the Russian platform openedu.ru were used for the analysis. It is well known that only a small percent of students who enrolled in MOOCs pass them through. Data mining methods could help to understand the causes of this problem. We tried to predict whether the student will finish an online course or not based on his results during the first weeks. Such prediction if it was performed early enough could help to keep students in the course.

Fazla değişken söz konusu olduğunda elle sınıflama yapmak zaman ve emek isteyen bir süreç haline gelmektedir. Böyle bir duruma örnek olan engelli bireylerin öz bakım aktivitelerinde yaşadıkları sorunlara göre sınıflandırılması, uzman terapistler için zaman alıcı bir süreçtir. Bu çalışmanın amacı uzman terapistlere zaman kazandırması açısından fiziksel ve motor engelli bireylerin öz bakım aktivitelerinde yaşadıkları sorunları tahmin edebilecek bir modelin geliştirilmesidir. Tahmin sürecinde yedi farklı (destek vektör makineleri, yapay sinir ağları, C5.0, CART, QUEST, CHAID, bayes ağları) veri madenciliği algoritmasından yararlanılmıştır. Söz konusu algoritmalar, tek olarak ve farklı kolektif modeller oluşturularak uygulanmıştır. Tek ve kolektif olarak uygulanan modellerin deneysel sonuçları kıyaslandığında, iki veya daha fazla algoritmanın birleştirilmesi ile uygulanan kolektif öğrenme tekniğinin tahmin performansını yükselttiği görülmüştür., In the case of too many variables, manual classification becomes a time- and labor-intensive process. For instance, classifying people with disabilities according to the problems they experience in self-care activities is a time-consuming process for specialist therapists. The aim of this study is to develop a model that can predict the problems experienced by physically and motor disabled individuals in self-care activities in order to reduce time spent by specialist therapist. Seven different data mining algorithms (support vector machines, artificial neural networks, C5.0, CART, QUEST, CHAID, bayesian networks) have been used in the estimation process. These algorithms have been applied individually and by forming different ensemble models. When experimental results of single and ensemble models were compared, it was seen that ensemble learning technique combined with two or more algorithms increased predictive performance.

The rapid development of network information technology has led to the explosive growth of global information data volume. People’s lives are more and more full of massive data. This era is vividly called “big data era”. With the application of big data technology, colleges and universities can predict the future by analyzing and mining the laws hidden behind the data, so that teachers and managers can improve their teaching and management decisions.

In today’s era, the effective use of data technology can not only benefit the development of enterprises, but also help to promote the progress and development of society. This requires enterprise technical personnel to strengthen the understanding of data technology, in order to make better use of data technology and play the role of data mining technology. This paper expounds the data mining technology and discusses its application in e-commerce, hoping to provide reference for relevant enterprises.