Handling Class Imbalance in Online Transaction Fraud Detection

With the rise of internet facilities, a greater number of people have started doing online transactions at an exponential rate in recent years as the online transaction system has eliminated the need of going to the bank physically for every transaction. However, the fraud cases have also increased causing the loss of money to the consumers. Hence, an effective fraud detection system is the need of the hour which can detect fraudulent transactions automatically in real-time. Generally, the genuine transactions are large in number than the fraudulent transactions which leads to the class imbalance problem. In this research work, an online transaction fraud detection system using deep learning has been proposed which can handle class imbalance problem by applying algorithm-level methods which modify the learning of the model to focus more on the minority class i.e., fraud transactions. A novel loss function named Weighted Hard- Reduced Focal Loss (WH-RFL) has been proposed which has achieved maximum fraud detection rate i.e., True Positive Rate (TPR) at the cost of misclassification of few genuine transactions as high TPR is preferred over a high True Negative Rate (TNR) in fraud detection system and same has been demonstrated using three publicly available imbalanced transactional datasets. Also, Thresholding has been applied to optimize the decision threshold using cross-validation to detect maximum number of frauds and it has been demonstrated by the experimental results that the selection of the right thresholding method with deep learning yields better results.     

Adaptive control of a hybrid electric vehicle

A decentralized adaptive control system (ACS) for a four motor-generator four-wheel drive hybrid electric vehicle (HEV) is designed and its ability to deal with unknown tire dynamics, changing road surfaces, and vehicle loading is evaluated. A system composed of four separate adaptive controllers is designed to control the vehicle's speed, steering, side slip, and energy management system. A nonlinear simulation model for the vehicle dynamics and its power train components is developed and used to evaluate the performance of the ACS, while the vehicle is simultaneously turning and accelerating or braking under varying loading and icing conditions.     

Sustainability Assessment of Urban Water Distribution Systems

A methodology is presented for determining sustainability indices for pressure and water age in water distribution systems (WDSs). These sustainability indices are based upon performance criteria including reliability, resiliency, and vulnerability. Pressure and water age are determined for a WDS as a function of operation time using the U.S. Environmental Protection Agency EPANET model. The values of pressure and water age are used to determine reliability, resiliency, and vulnerability performance criteria, which are then combined into the nodal sustainability indices for water age and pressure. In addition, the sustainability index (SI) computations are performed for zones to define the SI for water age and SI for pressure. A combined SI calculation is performed to produce an overall sustainability score for the entire zone in the water distribution network. The proposed methodology can be used to monitor the sustainability of existing WDSs and to help define alternative solutions including changes in pump operation and modifications to WDS to increase the sustainability.     

Non-derivable itemsets for fast outlier detection in large high-dimensional categorical data

Detecting outliers in a dataset is an important data mining task with many applications, such as detection of credit card fraud or network intrusions. Traditional methods assume numerical data and compute pair-wise distances among points. Recently, outlier detection methods were proposed for categorical and mixed-attribute data using the concept of Frequent Itemsets (FIs). These methods face challenges when dealing with large high-dimensional data, where the number of generated FIs can be extremely large. To address this issue, we propose several outlier detection schemes inspired by the well-known condensed representation of FIs, Non-Derivable Itemsets (NDIs). Specifically, we contrast a method based on frequent NDIs, FNDI-OD, and a method based on the negative border of NDIs, NBNDI-OD, with their previously proposed FI-based counterparts. We also explore outlier detection based on Non-Almost Derivable Itemsets (NADIs), which approximate the NDIs in the data given a δ parameter. Our proposed methods use a far smaller collection of sets than the FI collection in order to compute an anomaly score for each data point. Experiments on real-life data show that, as expected, methods based on NDIs and NADIs offer substantial advantages in terms of speed and scalability over FI-based Outlier Detection method. What is significant is that NDI-based methods exhibit similar or better detection accuracy compared to the FI-based methods, which supports our claim that the NDI representation is especially well-suited for the task of detecting outliers. At the same time, the NDI approximation scheme, NADIs is shown to exhibit similar accuracy to the NDI-based method for various δ values and further runtime performance gains. Finally, we offer an in-depth discussion and experimentation regarding the trade-offs of the proposed algorithms and the choice of parameter values.     

Conditional diagnosability of hypercubes under the comparison diagnosis model

Processor fault diagnosis plays an important role in multiprocessor systems for reliable computing, and the diagnosability of many well-known networks has been explored. Lai et al. proposed a novel measure of diagnosability, called conditional diagnosability, by adding an additional condition that any faulty set cannot contain all the neighbors of any vertex in a system. We make a contribution to the evaluation of diagnosability for hypercube networks under the comparison model and prove that the conditional diagnosability of n-dimensional Hypercube Q_n is 3(n ? 2) + 1 for n ? 5. The conditional diagnosability of Q_n is about three times larger than the classical diagnosability of Q_n.     

Disaster and Preparedness: Lessons from Hurricane Rita

This paper investigates the lessons learned and preparedness behaviours of businesses in the southeast Texas region affected by Hurricane Rita. The data were collected through an e‐mail survey sent to businesses in the southeast Texas region affected by Hurricane Rita. Findings from the study indicate that the majority of businesses took only ‘few’ or ‘some’ preparedness measures before the hurricane. However, those businesses that experienced impacts and losses due to Hurricane Rita report that they are now taking greater preparedness measures. This study identifies specific areas that should be addressed in a disaster preparedness plan based on information gathered from organizations that went through a major disaster first‐hand.     

Robustness of connectionist swimming controllers against random variation in neural connections

The ability to achieve high swimming speed and efficiency is very important to both the real lamprey and its robotic implementation. In previous studies, we used evolutionary algorithms to evolve biologically plausible connectionist swimming controllers for a simulated lamprey. This letter investigates the robustness and optimality of the best-evolved controllers as well as the biological controller hand-crafted by Ekeberg. Comparing cases of random variation in intrasegmental or intersegmental weights against each controller allows estimates of robustness to be made. We conduct experiments on the controllers' robustness at the excitation level, which corresponds to either the maximum swimming speed or efficiency by randomly varying the segmental connection weights and on some occasions also the intersegmental couplings, through varying noise ranges. Interestingly, although the swimming performance (in terms of maximum speed and efficiency) of the Ekeberg biological controller is not as good as that of the artificially evolved controllers, it is relatively robust against noise in the neural networks. This suggests that the natural evolutions have evolved a swimming controller that is good enough to survive in the real world. Our findings could inspire neurobiologists to conduct real physiological experiments to gain a better understanding on how neural connectivity affects behavior. The results can also be applied to control an artificial lamprey in simulation and possibly also a robotic one.     

Gastrointestinal fluids proteomics

Seventy million people suffer from diseases of the gastrointestinal tract annually in US, translating to US$85.5 billion in direct healthcare costs. The debilitating effects of these gastrointestinal (GI) diseases can be circumvented with good biomarkers for early detection of these disorders, which will greatly increase the success of curative treatments. GI fluids represent a potential reservoir of biomarkers for early diagnosis of various GI and systemic diseases since these fluids are the most proximal fluid bathing diseased cells. They are anticipated to have proteomes that closely reflect the ensemble of proteins secreted from the respective GI tissues. Most importantly, the disease markers present in GI fluids should be present in higher concentrations than in sera, thus offering greater sensitivity in their detection. However, proteome analysis of GI fluids can be complex mainly due to the dynamic range of protein content and the numerous PTMs of proteins in each specialized GI compartment. This review attempts to discuss the physiology of the various GI fluids, the special technical considerations required for proteome analysis of each fluid, as well as to summarize the current state of knowledge of biomarker discoveries and clinical utility of GI fluids such as salivary, gastric, pancreatic, and biliary secretions.