‎k-means‎

Digital image segmentation plays an important role in noise reduction and pixel clustering for pre-processing of deep learning or feature extraction. The classic Self-Organizing Map (SOM) algorithm is a well-known unsupervised clustering neural network model. This classic method works on continuous data instead of discrete data sets with a widely scattered distribution. The novel SOM(SOM2) modelling solved this problem for the classic, simple tabular discrete data set but not for the digital image data. As the essence of digital image pixels data are different from tabular datasets, we have to look at them differently. This paper proposes exploiting the novel SOM method with a hybrid combination of the fuzzy C-Means and K-means convolution filter as image segmentation and noise reduction with soft and hard segmentation as entropy reduction for natural digital images. The main approach of this paper is the segmentation of image contents for the reduction of noises and saturation pixels by entropy criteria. Based on the resulting paper, the combination of SOM2 with FCM for soft segmentation 47%-and the combination of SOM2 with k-means convolution for hard segmentation 33% can reduce the entropy of the original image on average.

Given the importance of policyholder classification in helping to make a good decision in predicting optimal premiums for actuaries.This paper proposes, first, an optimal construction of policyholder classes. Second, Poisson-negative Binomial mixture regression model is proposed as an alternative to deal with the overdispersion of these classes.The proposed method is unique in that it takes Tunisian data and classifies the insured population based on the K-means approach which is an unsupervised machine learning algorithm. The choice of the model becomes extremely difficult due to the presence of zero mass in one of the classes and the significant degree of overdispersion. For this purpose, we proposed a mixture regression model that leads us to estimate the density of each class and to predict its probability distribution that allows us to understand the underlying properties of our data. In the learning phase, we estimate the values of the model parameters using the Expectation-Maximization algorithm. This allows us to determine the probability of occurrence of each new insured to create the most accurate classification. The goal of using mixed regression is to get as heterogeneous a classification as possible while having a better approximation. The proposed mixed regression model, which uses a number of factors, has been evaluated on different criteria, including mean square error, variance, chi-square test and accuracy. According to the experimental findings on several datasets, the approach can reach an overall accuracy of 80%. Then, the application on real Tunisian data shows the effectiveness of using the mixed regression model.

One of well-known techniques in data mining is clustering. Clustering method which is very popular is K-means cluster because its algorithm is very easy and simple. However, K-means cluster has some weaknesses, one of which is that the cluster result is sensitive towards centroid initialization so that the cluster result tends to local optimal. This paper explains the modification of K-means cluster, that is, K-means hybridization with ant colony optimization (K-ACO). Ant Colony Optimization (ACO) is optimization algorithm based on ant colony behavior. Through K-ACO, the weaknesses of cluster result which tends to local optimal can be overcome well. The application of hybrid method of K-ACO with the use of R program gives better accuracy compared to K-means cluster. K-means cluster accuracy yielded by Minitab, Mathlab, and SAS at iris data is 89%. Meanwhile, K-ACO hybrid clustering with R program simulated on 38 treatments with 3-time repetitions gives accuracy result of 93,10%.

Human Activity Recognition (HAR) systems used in healthcare have attracted much attention in recent years. A HAR system consists of a wearable device with sensors. HAR has been used to suggest several machine learning (ML) algorithms. However, only a few research have looked at how to evaluate HAR to identify physical activities. Nevertheless, obtaining an explanation for their performances is complicated by two factors: the lack of implementation specifics and the lack of a baseline evaluation setup that makes comparisons unfair. For establishing effective and efficient ML–HAR of computers and networks, this study uses ten common unsupervised and supervised ML algorithms. The decision tree (DT), artificial neural network (ANN), naive Bayes (NB), k-nearest neighbor (k-NN), support vector machine (SVM), random forest (RF), and XGBoost (XGB) algorithms are among the supervised ML algorithms, while the k-means, expectation-maximization (EM), and self-organizing maps (SOM) algorithms are among the unsupervised ML algorithms. Multiple algorithms models are presented, and the turning and training parameters in ML (DT, ANN, NB, KNN, SVM, RF, XGB) of each method are investigated in order to obtain the best classifier assessment. Differ from earlier research, this research measures the true negative and positive rates, precision, accuracy, F-Score as well as recall of 81 ML-HAR models to assess their performance. Because time complexity is a significant element in HAR, the ML-HAR models training and testing time are also taken into account when evaluating their performance efficiency. The mobile health care (M\_HEALTH CARE) dataset, which includes real-world network activity, is used to test the ML-HAR models. In general, the XGB outperforms the DT-HAR, k-NN-HAR, and NB-HAR models in recognizing human activities, with recall, precision, and f-scores of 0.99, 0.99, and 0.99 for each, respectively, for health care mobile recognition.

Concerning life insurance, penetration rate is one of the main goal of every developed insurance industry. In this sense systematic marketing is a significant component in strategic plan of insurance companies. To achieve the goal insurers need to group their client into different groups in which some common features are shared and people demonstrate a similar pattern. This paper utilizes K-means clustering as an unsupervised learning algorithm in order to divide customers into number of clusters. The clusters are constructed based on two independent variables namely; car and life insurance premiums. Then the descriptive statistics of other determining features are provided with which the most willing group in purchasing life insurance is presented.

‎In some fields‎, ‎there is an interest in distinguishing different geometrical objects from each other‎. ‎A field of research that studies the objects from a statistical point of view‎, ‎provided they are‎ ‎invariant under translation‎, ‎rotation and scaling effects‎, ‎is known as the statistical shape analysis‎. ‎Having some objects that are registered using key points on the outline of the objects‎, ‎the main purpose‎ ‎of this paper is to compare two popular clustering procedures to cluster objects‎. ‎We also use some indexes‎ ‎to evaluate our clustering application‎. ‎The proposed methods are applied to the real life data.