Search

Hospital cost analysis (HCA) becomes a key topic and forefront of politics, social welfare and medical discourse. HCA includes a wide range of expenses; yet the foremost attention relates to the money expense in which hospital managers would like to draw a figure of incomes in the past and future. Based on the HCA results, they can develop many plans for improving hospital's service quality and investing in potential healthcare services in order to deliver better services with lower costs. Machine learning methods are often opted for prediction in HCA. In this paper, we propose a new method for HCA that uses genetic algorithm (GA) and artificial neural network (ANN). Operators of GA are used to boost up calculation to get optimal weights in the forward propagation of ANN. Experiments on a real database of Hanoi Medical University Hospital (HMUH) including calculus of kidney and ureter inpatients show that the new method achieves better accuracy than the relevant ones including linear regression, K-nearest neighbors (KNN), ANN and deep learning. The mean squared error of the proposed model gets the lowest value (0.00360), compared to those of deep learning, KNN and linear regression which are 0.00901, 0.01205 and 0.01718 respectively. [ABSTRACT FROM AUTHOR]

Fuzzy minimum spanning tree (FMST) has emerged from various real-life applications in different areas by considering uncertainty that exists in arc lengths of a fuzzy graph. In most relevant studies regarding FMST, type 1 fuzzy set was used to represent edge weights. Nonetheless, its membership values are totally crisp which is hard to determine its exact value by human perception. Interval type 2 fuzzy set (IT2FS) increases the number of degrees of freedom to express uncertainty of the edge weight and has more capacity to describe fuzzy information in a logically correct manner. In this paper, we propose the minimum spanning tree problem with undirected connected weighted interval type 2 fuzzy graph (FMST-IT2FS). Herein, the interval type 2 fuzzy set is used to represent the arc lengths of a fuzzy graph. Then, a new genetic algorithm is proposed to solve the FMST-IT2FS problem with the addition, ranking and defuzzification of IT2FSs being used. Illustrative examples are included to demonstrate the effectiveness of the proposed algorithm. [ABSTRACT FROM AUTHOR]

Artificial neural network (ANN) has been well applied in pattern recognition, classification and machine learning thanks to its high performance. Most ANNs are designed by a static structure whose weights are trained during a learning process by supervised or unsupervised methods. These training methods require a set of initial weights values, which are normally randomly generated, with different initial sets of weight values leading to different convergent ANNs for the same training set. Dealing with these drawbacks, a trend of dynamic ANN was invoked in the past year. However, they are either too complex or far from practical applications such as in the pathology predictor in binary multi-input multi-output (MIMO) problems, when the role of a symptom is considered as an agent, a pathology predictor's outcome is formed by action of active agents while other agents' activities seem to be ignored or have mirror effects. In this paper, we propose a new dynamic structural ANN for MIMO problems based on the dependency graph, which gives clear cause and result relationships between inputs and outputs. The new ANN has the dynamic structure of hidden layer as a directed graph showing the relation between input, hidden and output nodes. The properties of the new dynamic structural ANN are experienced with a pathology problem and its learning methods' performances are compared on a real well known dataset. The result shows that both approaches for structural learning process improve the quality of ANNs during learning iteration. [ABSTRACT FROM AUTHOR]

In the original article, family name of one of the co-authors (Duong Thi Thu Huyen) has been missed in the online publication. [ABSTRACT FROM AUTHOR]

A genetic algorithm (GA) belongs to the class of evolutionary algorithms and it is one of the most studied heuristic algorithms to solve graph coloring problems. In this paper, we propose a new GA algorithm for the total graph coloring problem. To the best of our knowledge, no algorithm based on a GA exists in the literature for total graph coloring. In the proposed approach, a novel encoding scheme is introduced, where all the edges and vertices of the graph are represented in a chromosome without any repetition. For the initialization of the population, a greedy algorithm is used to determine the total number of colors required for a total coloring of the graph. The number of colors is used as the fitness value of a chromosome which depends on the sequence of vertices and edges representing the chromosome. We introduce a convergence criteria for GA based on the total coloring conjecture. A two-point crossover and mutation operations, suitable for total coloring, are suggested. The proposed algorithm is applied on some well-known and standard graphs. In our computational tests, graphs are used with a maximum number of 690 vertices and 6650 edges of the graph, respectively. The proposed algorithm determines an optimal solution for 21 graphs among the 27 graphs. The solution of remaining the 6 graphs is near optimal and differs by at most one unit from the optimal value. The results show the effectiveness of the proposed approach. [ABSTRACT FROM AUTHOR]

Image steganography plays a vital role in hiding significant secret information within any input image. Numerous steganography techniques have been carried out to hide the secret information in images. In the current scenario, frequency domain techniques are widely preferred which are mostly transform based approaches. Conventionally, the secret data is hidden randomly in the coefficients of the transformed image. However, such random data hiding techniques lead to inferior performance of the overall system. Thus, using an optimization algorithm is obligatory to find out the optimal coefficients. Genetic Algorithm (GA) is normally used for selecting the optimal transform coefficients to enhance the system performance. However, conventional GA based approaches are highly random in nature which again leads to inaccurate results. In this work, a modified GA approach was proposed to determine the optimal coefficients in order to improve the embedding capacity and stego image quality. The achieved average peak signal to noise ratio (PSNR) was 50.29 dB with embedding capacity of 139361 bits. These experimental results validate the practical feasibility of the proposed methodology for image steganography. [ABSTRACT FROM AUTHOR]