Attribute weighted Naive Bayes classifier using a local optimization

The Naive Bayes classifier is a popular classification technique for data mining and machine learning. It has been shown to be very effective on a variety of data classification problems. However, the strong assumption that all attributes are conditionally independent given the class is often violated in real-world applications. Numerous methods have been proposed in order to improve the performance of the Naive Bayes classifier by alleviating the attribute independence assumption. However, violation of the independence assumption can increase the expected error. Another alternative is assigning the weights for attributes. In this paper, we propose a novel attribute weighted Naive Bayes classifier by considering weights to the conditional probabilities. An objective function is modeled and taken into account, which is based on the structure of the Naive Bayes classifier and the attribute weights. The optimal weights are determined by a local optimization method using the quasisecant method. In the proposed approach, the Naive Bayes classifier is taken as a starting point. We report the results of numerical experiments on several real-world data sets in binary classification, which show the efficiency of the proposed method.     

Estimation of the critical properties of compounds using volume-based thermodynamics

The importance, yet scarcity of the critical constants of thermally unstable fluids warrants the development of reliable methods for the estimation of these essential thermodynamic properties. A thorough investigation undertaken in this study on 1,589 compounds belonging to 83 chemical classes, indicated that the ratio of critical temperature to critical pressure of both low and high molecular weight compounds could be well expressed in terms of their volumetric properties. In addition, two new methodologies are presented for estimating V_c, as well as an indirect approach for prediction of T_c from surface tension data, altogether allowing the calculation of Z_c. Moreover, comparative studies are made with five group contribution methods. It is also demonstrated that by employing the Peng-Robinson EOS, and without prior knowledge of the critical properties, it is possible to calculate various thermophysical properties including P^sat., T_b, , ∆H^vap. , C_p, and even the T_c and P_c themselves.     

The Impact of the Microbiome on Resistance to Cancer Treatment with Chemotherapeutic Agents and Immunotherapy

Understanding the mechanisms of resistance to therapy in human cancer cells has become a multifaceted limiting factor to achieving optimal cures in cancer patients. Besides genetic and epigenetic alterations, enhanced DNA damage repair activity, deregulation of cell death, overexpression of transmembrane transporters, and complex interactions within the tumor microenvironment, other mechanisms of cancer treatment resistance have been recently proposed. In this review, we will summarize the preclinical and clinical studies highlighting the critical role of the microbiome in the efficacy of cancer treatment, concerning mainly chemotherapy and immunotherapy with immune checkpoint inhibitors. In addition to involvement in drug metabolism and immune surveillance, the production of microbiota-derived metabolites might represent the link between gut/intratumoral bacteria and response to anticancer therapies. Importantly, an emerging trend of using microbiota modulation by probiotics and fecal microbiota transplantation (FMT) to overcome cancer treatment resistance will be also discussed.     

Effect of calcinations of starting powder on mechanical properties of hydroxyapatite–alumina bioceramic composite

The effect of calcinations of starting powder on the mechanical properties of hydroxyapatite (HA)-based bioceramic composite was investigated. The calcinations of HA powder in air at 900 °C increased the crystallinity as well as the size of the powder. Ball milling after the calcinations was effective in eliminating large agglomerates in the powder. When the powder was mixed with reinforcing Al₂O₃ powder, the mixture became fine and homogeneous. The flexural strength of HA–Al₂O₃ composite was increased by the calcinations processes at all the Al₂O₃ concentration. However, the fracture toughness was not much influenced by the calcinations. These results lead to the conclusion that the calcinations process effectively reduced the critical flaw size in the body by removing the agglomerates in the HA powder.     

An analysis for AlGaN/GaN modulation doped field effect transistor using accurate velocity-field dependence for high power microwave frequency applications

An analytical model of Al_0.15Ga_0.85N/GaN modulation doped field effect transistor (MODFET), which uses an accurate velocity field relationship and incorporates the dominant effect of piezoelectric polarization induced charge at the AlGaN/GaN interface is presented. The effect of traps has also been taken into account. The calculated DC characteristics are in excellent agreement with the measured data. The model is extended to predict the microwave performance of the device. High current levels (>500 mA/mm), large transconductance (160.83 mS/mm) and a high cutoff frequency (9.6 GHz) have been achieved analytically and are in close agreement with the experimental data.     

Multisource multicasting in IP/MPLS over WDM networks

Multicasting applications such as multimedia conferencing, online multiplayer interactive games, and distance learning are becoming increasingly popular. With multiprotocol label switching, Internet protocol networks can offer quality of service and traffic engineering capabilities. This article introduces several approaches for multisource multicast sessions in the context of IP over WDM networks and evaluates their performance in terms of blocking probability, time complexity, and memory consumption. Our simulation study shows that among all the approaches, the newly proposed approach, known as one Bidirectional Tree with Just enough bandwidth reserved on each link of the tree, achieves the best overall performance.