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.     

Automatic Heart Disease Detection by Classification of Ventricular Arrhythmias on ECG Using Machine Learning

This paper focuses on detecting diseased signals and arrhythmias classification into two classes: ventricular tachycardia and premature ventricular contraction. The sole purpose of the signal detection is used to determine if a signal has been collected from a healthy or sick person. The proposed research approach presents a mathematical model for the signal detector based on calculating the instantaneous frequency (IF). Once a signal taken from a patient is detected, then the classifier takes that signal as input and classifies the target disease by predicting the class label. While applying the classifier, templates are designed separately for ventricular tachycardia and premature ventricular contraction. Similarities of a given signal with both the templates are computed in the spectral domain. The empirical analysis reveals precisions for the detector and the applied classifier are 100% and 77.27%, respectively. Moreover, instantaneous frequency analysis provides a benchmark that IF of a normal signal ranges from 0.8 to 1.1 Hz whereas IF range for ventricular tachycardia and premature ventricular contraction is 0.08–0.6 Hz. This indicates a serious loss of high-frequency contents in the spectrum, implying that the heart’s overall activity is slowed down. This study may help medical practitioners in detecting the heart disease type based on signal analysis.     

A Hybrid Duo-Deep Learning and Best Features Based Framework forActionRecognition

Human Action Recognition (HAR) is a current research topic in the field of computer vision that is based on an important application known as video surveillance. Researchers in computer vision have introduced various intelligent methods based on deep learning and machine learning, but they still face many challenges such as similarity in various actions and redundant features. We proposed a framework for accurate human action recognition (HAR) based on deep learning and an improved features optimization algorithm in this paper. From deep learning feature extraction to feature classification, the proposed framework includes several critical steps. Before training fine-tuned deep learning models – MobileNet-V2 and Darknet53 – the original video frames are normalized. For feature extraction, pre-trained deep models are used, which are fused using the canonical correlation approach. Following that, an improved particle swarm optimization (IPSO)-based algorithm is used to select the best features. Following that, the selected features were used to classify actions using various classifiers. The experimental process was performed on six publicly available datasets such as KTH, UT-Interaction, UCF Sports, Hollywood, IXMAS, and UCF YouTube, which attained an accuracy of 98.3%, 98.9%, 99.8%, 99.6%, 98.6%, and 100%, respectively. In comparison with existing techniques, it is observed that the proposed framework achieved improved accuracy.     

Identifying factors that increase the likelihood of driving after drinking among college students

Driving after drinking (DAD) is a serious public health concern found to be more common among college students than those of other age groups or same-aged non-college peers. The current study examined potential predictors of DAD among a dual-site sample of 3753 (65% female, 58% Caucasian) college students. Results showed that 19.1% of respondents had driven after 3 or more drinks and 8.6% had driven after 5 or more drinks in the past 3 months. A logistic regression model showed that male status, fraternity or sorority affiliation, family history of alcohol abuse, medium or heavy drinking (as compared to light drinking), more approving self-attitudes toward DAD, and alcohol expectancies for sexual enhancement and risk/aggression were independently associated with driving after drinking over and above covariates. These results extend the current understanding of this high risk drinking behavior in collegiate populations and provide implications for preventive strategies. Findings indicate that in addition to targeting at-risk subgroups, valuable directions for DAD-related interventions may include focusing on lowering both self-approval of DAD and alcohol-related expectancies, particularly those associated with risk/aggression and sexuality.     

In situ biosensing with a surface plasmon resonance fiber grating aptasensor

Surface plasmon resonance (SPR) biosensors prepared using optical fibers can be used as a cost-effective and relatively simple-to-implement alternative to well established biosensor platforms for monitoring biomolecular interactions in situ or possibly in vivo. The fiber biosensor presented in this study utilizes an in-fiber tilted Bragg grating to excite the SPR on the surface of the sensor over a large range of external medium refractive indices, with minimal cross-sensitivity to temperature and without compromising the structural integrity of the fiber. The label-free biorecognition scheme used demonstrates that the sensor relies on the functionalization of the gold-coated fiber with aptamers, synthetic DNA sequences that bind with high specificity to a given target. In addition to monitoring the functionalization of the fiber by the aptamers in real-time, the results also show how the fiber biosensor can detect the presence of the aptamer's target, in various concentrations of thrombin in buffer and serum solutions. The findings also show how the SPR biosensor can be used to evaluate the dissociation constant (K(d)), as the binding constant agrees with values already reported in the literature.     

Crystalline ZnO and ZnO/TiO2 nanoparticles derived from tert-butyl N-(2 mercaptoethyl)carbamatozinc(II) chelate: Electrocatalytic studies for H2 generation in alkaline electrolytes

A newly synthesized zinc(II) complex, namely tert-butyl N-(2 mercaptoethyl)carbamatozinc(II) complex [Zn(Boc-S)₂] (Boc = tert-butyl N-[2-mercaptoethyl]carbamate), has been used as an organozinc precursor for the production of crystalline ZnO and ZnO/TiO₂ nanoparticles. The synthesized complex and the obtained nanomaterials were fully characterized using various spectroscopic and surface analysis techniques. Their surface morphology, chemical purity and stoichiometry have been investigated by scanning electron microscopy (SEM), energy dispersive X-ray analysis (EDX) as well as X-ray fluorescence. The synthesized Zn(II) molecular complex, ZnO and ZnO/TiO₂ nanomaterials have been tested in alkaline aqueous solution (1.0 MNaOH) for the hydrogen evolution reaction (HER) using various electrochemical techniques. The results revealed high HER catalytic performance of ZnO and ZnO/TiO₂ cathode materials, with the latter exhibiting higher catalytic activity recording an exchange current density (j_o) of 0.3 mA cm⁻². This current value, which approaches that of Pt wire (0.5 mA cm⁻²), cross-sectional area ~0.008 cm², is about 11 and 100 times greater than those measured for ZnO alone (0.028 mA cm⁻²) and TiO₂ alone (0.0032 mA cm⁻²), respectively. Moderate catalytic activity was recorded for the complex catalyst, namely GC-Zn(Boc-S)₂ with j_o value of (0.01 mA cm⁻²). Tafel slope values of 130 and 122 mV dec⁻¹ were calculated for ZnO and ZnO/TiO₂, respectively. Such Tafel slope values, which are close to that of the Pt wire (120 mV dec⁻¹), referred to a Volmer-controlled HER kinetics. Other important electrochemical parameters describing the kinetics of the HER, such as roughness factor (R_f) and turnover frequency (TOF) were also estimated and discussed. The high numerical values of the various HER kinetic parameters recorded for the ZnO/TiO₂ catalyst, in addition to its high stability and durability (stable for up to 10 000 continuous cathodic polarization cycles), besides maintaining its morphology and chemical composition after stability test (confirmed from SEM/EDX and XRD examinations), located it in a privileged position among the most efficient HER electrocatalysts reported in the literature.     

Study of oxygen plasma pre-treatment of polyester fabric for improved polypyrrole adhesion

In this work, we have systematically studied the improvement of binding of polypyrrole with polyethylene terephthalate (PET) thin films and fabrics using low pressure oxygen plasma. A range of plasma treatment times were employed to investigate plasma induced effects on surface roughness, surface chemistry and hydrophilicity. Modifications of PET films were studied with respect to surface morphology by means of atomic force and scanning electron microscopy. Chemical effects of plasma treatment were studied using X-ray photoelectron spectroscopy. Results showed that both the increase in surface functionalisation by carboxylic groups and formation of nano size roughness contributed to improved adhesion and conductivity.     

Interface Immobilization Chemistry of cRGD‐based Peptides Regulates Integrin Mediated Cell Adhesion

The interaction of specific surface receptors of the integrin family with different extracellular matrix‐based ligands is of utmost importance for the cellular adhesion process. A ligand consists of an integrin‐binding group, here cyclic RGDfX, a spacer molecule that lifts the integrin‐binding group from the surface and a surface anchoring group. c(‐RGDfX‐) peptides are bound to gold nanoparticle structured surfaces via polyproline, polyethylene glycol or aminohexanoic acid containing spacers of different lengths. Although keeping the integrin‐binding c(‐RGDfX‐) peptides constant for all compounds, changes of the ligand's spacer chemistry and length reveal significant differences in cell adhesion activation and focal adhesion formation. Polyproline‐based peptides demonstrate improved cell adhesion kinetics and focal adhesion formation compared with common aminohexanoic acid or polyethylene glycol spacers. Binding activity can additionally be improved by applying ligands with two head groups, inducing a multimeric effect. This study gives insights into spacer‐based differences in integrin‐driven cell adhesion processes and remarkably highlights the polyproline‐based spacers as suitable ligand‐presenting templates for surface functionalization.     

Sensitization to Alternaria and Cladosporium by the age of 4 years

Of 1218 children born on the Isle of Wight in 1989/90, and followed for atopy at age 4 years, 981 were skin-prick tested with a battery of allergens. Of these 61 (6%) reacted positively to Alternaria alternata and Cladosporium herbarum (47 to Alternaria, 21 to Cladosporium and seven to both). Twenty-four (39%) were asymptomatic (latent atopy) of which 12 had a single positive reaction either to Alternaria or Cladosporium. Asthma was the most common disease in children sensitized to moulds. Alternaria sensitization correlated positively with clinical diagnosis of asthma (P < 0.01), eczema (P < 0.001) and rhinitis (P < 0.05). Likewise, Cladosporium sensitivity correlated with a diagnoses of asthma, eczema and rhinitis (all P < 0.05). Age of the house correlated with reported damp and lack of central heating (both P < 0.001), but not with sensitization to moulds. An association between the presence of damp or age of the house and mould allergy was confounded by 21 children moving house in the first 4 years. Exposure to pets, passive tobacco smoking and season of birth had no bearing on mould sensitivity. At 4 years of age Alternaria and Cladosporium were the third most common causes of sensitization, i.e. after house dust mite and grass pollen.