Asian Journal of Engineering and Applied Technology

Utilizing Prediction Intervals for Unsupervised Detection of Fraudulent Transactions: A Case Study

Isuru Hewapathirana — 2022-10-28

Money laundering operations have a high negative impact on the growth of a country’s national economy. As all financial sectors are increasingly being integrated, it is vital to implement effective technological measures to address these fraudulent operations. Machine learning methods are widely used to classify an incoming transaction as fraudulent or non-fraudulent by analyzing the behaviour of past transactions. Unsupervised machine learning methods do not require label information on past transactions, and a classification is made solely based on the distribution of the transaction. This research presents three unsupervised classification methods: ordinary least squares regression-based (OLS) fraud detection, random forest-based (RF) fraud detection and dropout neural network-based (DNN) fraud detection. For each method, the goal is to classify an incoming transaction amount as fraudulent or non-fraudulent. The novelty in the proposed approach is the application of prediction interval calculation for automatically validating incoming transactions. The three methods are applied to a real-world dataset of credit card transactions. The fraud labels available for the dataset are removed during the model training phase but are later used to evaluate the performance of the final predictions. The performance of the proposed methods is further compared with two other unsupervised state-of-the-art methods. Based on the experimental results, the OLS and RF methods show the best performance in predicting the correct label of a transaction, while the DNN method is the most robust method for detecting fraudulent transactions. This novel concept of calculating prediction intervals for validating an incoming transaction introduces a new direction for unsupervised fraud detection. Since fraud labels on past transactions are not required for training, the proposed methods can be applied in an online setting to different areas, such as detecting money laundering activities, telecommunication fraud and intrusion detection.

Assessment of Surface Quality in Extrusion Honing Process Using Dimensional Analysis Approach

S. L. N. Jayasimha — 2022-11-10

It is critical to obtain the desired surface quality on the internal and external portions of machined part. The estimated level of surface texture can be induced on the exterior regions using traditional finishing processes such as grinding, honing, and so on. While the problem emerges when processing core miniature components such as micro bores, inlet/outlet valves etc. The EH process overcomes this limitation of conventional finishing method. It is a novel micro machining process that extrudes the pressured flow of carrier media blended with abrasives into the confined passage to generate desired level of surface texture. Owing to abrasion process micro machining occurs by taking away the negligible amount of stock material. The present study focuses on the impact of number of passes at the specimen’s entry and exit sides of the carrier media. The improvements in surface finish (R_a) on both side i.e., entry and exit are evaluated as well. A dimensionless expression for R_a is also developed. The relationship is implemented using Buckingham’s π theorem and comparison of developed model is performed with experimental results. SEM analysis is made to portray surface texture produced by selected process parameters such as number of passes, volume fraction and grit size of abrasive grains.

Acute Toxicity of Monocrotophos on Histological Alterations in the Anomuran Crab, Emerita asiatica (H. Milne Edwards, 1837)

J. Sivakumar — 2022-11-30

This study characterized the acute toxicity of Monocrotophos on histological alterations in the different organs like gills, Hepatopancreas and ovary of Anomuran Crab, Emerita asiatica. Though Emerita asiatica is not a commercially viable crab, but it plays a vital role in the coastal environment to maintain a stable marine ecosystem. Several steps and precautions measures should be taken to consume these members of the marine food chain to have aunwavering ecosystem and also to protect this species from extinction. Thus, it can be concluded that the use of Monocrotophos which has been legally banned in India is justified. It has been proved by several workers and has been conformed in present investigation that use of this organophosphate causes serious damage to the vital organ of sand crabs gill, hepatopancreas and ovary. The utilization of Monocrotophos should be minimized in the agricultural field area near to the coastal ecosystem.

Wavelength Assignment Techniques with Converters for All Optical Networks (AONs)

Paul Franklin — 2022-12-19

We present an optical routing and wavelength assignment (RWA) technique for the best placement of the wavelength converters. The wavelength converter placement was considered independently at the complete nodes and partial nodes. In comparison to existing approaches, our technique performs better and requires fewer wavelength converters to get the required performance. As much as 7%, the likelihood of blocking is decreased. The study also demonstrates that the top four nodes have the biggest influence over the blocking performance of the network. It is clear that placing the converters in the right location is just as crucial to improving performance as simply adding more converters.

Design and Fabrication of an Instrument for Splitted Jute Fiber Length Measurement

Md. Redwanul Islam — 2022-12-19

Fibers are the main raw materials in the textile industry. Now-a-days various types of fibers are available around the world, but all fibers are not textile fibers. Because for being textile fibers, the fibers must have some particular physical and chemical properties. The length of natural fiber plays an important role both in the spinning process and product quality. In this work, it is highlighted about a long staple natural fiber which is called jute fiber. In the 19th century, jute fiber was called the *Golden Fiber* of Bangladesh. After that this fiber lost its name and fame for the cause of our negligence and lack of knowledge. It is very difficult to know about jute because a few people were working on it. Though it’s a long fiber but there are some problems with its fixed length because this fiber consists of a large number of ultimate cells whose length is 1.5 to 4 millimeters, and these cells are attached by lignin which is known as natural cement. A long fiber which is 5 to 10 feet forms after attaching these ultimate cells. During the retting period some of the lignin layers break down and long fiber length of fibers becomes shorter. Another important thing is these fibers have a mesh structure for this reason it’s difficult to identify the actual length of the fibers. That’s why fibers obtained after retting are passed through jute carding machines. The main function of the jute carding machine is to split up and break down the mesh structure. After carding, it’s quite easy to measure the length of splitted jute fibers. There are not any suit-able machines for measuring the length of jute fibers. Previously people use the scale for measuring the jute fibers' length manually, through this it is so difficult to measure the length of the long fibers accurately. Here an instrument has been designed with Arduino UNO, Ultrasonic distance sensor, Image-j software, C-programming and fabricated for measuring the fibers' length easily and accurately as well as this instrument can measure the number of fibers in that particular length. By using this instrument anyone can easily measure the length of the fibers and the number of fibers within a short possible time efficiently.