Journal of Engineering and Technological Sciences
Not a member yet
888 research outputs found
Sort by
Gene Family Abundance Visualization based on Feature Selection Combined Deep Learning to Improve Disease Diagnosis
Get PDFAdvancements in machine learning in general and in deep learning in particular have achieved great success in numerous fields. For personalized medicine approaches, frameworks derived from learning algorithms play an important role in supporting scientists to investigate and explore novel data sources such as metagenomic data to develop and examine methodologies to improve human healthcare. Some challenges when processing this data type include its very high dimensionality and the complexity of diseases. Metagenomic data that include gene families often have millions of features. This leads to a further increase of complexity in processing and requires a huge amount of time for computation. In this study, we propose a method combining feature selection using perceptron weight-based filters and synthetic image generation to leverage deep-learning advancements in order to predict various diseases based on gene family abundance data. An experiment was conducted using gene family datasets of five diseases, i.e. liver cirrhosis, obesity, inflammatory bowel diseases, type 2 diabetes, and colorectal cancer. The proposed method provides not only visualization for gene family abundance data but also achieved a promising performance level
On Minimal Second-order IIR Bandpass Filters with Constrained Poles and Zeros
Get PDFIn this paper, several forms of infinite impulse response (IIR) bandpass filters with constrained poles and zeros are presented and compared. The comparison includes the filter structure, the frequency ranges and a number of controlled parameters that affect computational efforts. Using the relationship between bandpass and notch filters, the two presented filters were originally developed for notch filters. This paper also proposes a second-order IIR bandpass filter structure that constrains poles and zeros and can be used as a minimal parameter adaptive digital second-order filter. The proposed filter has a wider frequency range and more flexibility in the range values of the adaptation parameters
Finite Element Simulation of Vacuum Preloading at Palembang – Indralaya Toll Project
Get PDFMethods for the prediction of soil behavior during the application of vacuum preloading are available but have not been used precisely and have not been proven yet in Indonesia. There are two common approaches to vacuum preloading simulation, based on the application of a uniform external load to the vacuum area, and based on suddenly lowering the groundwater level to create vacuum conditions, respectively. This affects the settlement, lateral deformation, and pore pressure predictions. The objective of this research was to improve the prediction of soil behavior based on high-quality field data by using state of the art vacuum preloading simulations. The results were compared with those of a series of instrumentation equipment, i.e. a settlement plate, an extensometer, and a piezometer. This research used data from the Palembang-Indralaya Toll Road, a section of the Trans Sumatera Toll Road that is approximately 22 km long and has an embankment height of about 4 m to 9 m. It was built over a swampy soft soil area, using vacuum preloading to improve the soil. Axisymmetric analysis of vacuum preloading was conducted for a single-drain system, plane-strain analysis was conducted for single- and multiple-drain systems, and 3D analysis was conducted for single-drain, multiple-drain, and cluster-drain systems. The results show that the proposed method produced a good correlation between the predicted data and the recorded monitoring data
Airflow Characteristics Investigation of a Diesel Engine for Different Helical Port Openings and Engine Speeds
Get PDFIntake airflow characteristics are essential for the performance of diesel engines. However, previous investigations of these airflow characteristics were mostly performed on two-valve engines despite the difference between the airflow of two-valve and four-valve engines. Therefore, in this study, particle image velocimetry (PIV) investigations were performed on a four-valve diesel engine. The investigations were conducted under different engine speeds and helical port openings using a swirl control valve (SCV). The results suggest that the position of the swirl center does not significantly shift with different engine speeds and helical port openings, as the dynamics of the flow remained closely similar. The trends of the airflow characteristics can be best observed during the compression stroke. A higher engine speed increases the angular velocity of the engine more compared to the increase of the airflow velocity and results in a lower swirl ratio of the flow. On the other hand, a higher engine speed leads to a higher mean velocity and the variation of velocity results in a larger turbulence intensity of the flow. Increasing the helical port opening brings a reduction in the swirl ratio and turbulence intensity as more airflow from the helical port disturbs the airflow from the tangential port
Performance Investigation of MIMO Based CO-OFDM FSO Communication Link for BPSK, QPSK and 16-QAM under the Influence of Reed Solomon Codes
Get PDFThe MIMO based CO-OFDM FSO communication system is emerging as a promising approach to meet the future bandwidth requirements for seamless communication. The atmosphere being the propagation medium is a major hindrance in wide-scale acceptability of FSO technology. For seamless and error-free transmission and reception of data, a novel concept of MIMO integrated with RS code is proposed in this paper. The system performance of an RS 64 (RS (255,127)) coded MIMO-based CO-OFDM FSO communication link was investigated using BPSK, QPSK and 16-QAM under the combined effects of geometric losses, path losses and atmospheric attenuations at a hitherto un-investigated data rate of 40 Gbps and a link distance of 5 km. The modified gamma-gamma distribution was used for modeling a moderately turbulent channel. With link length varying over a range of 1 to 5 km, error correction was maximum in 16-QAM as compared to BPSK and QPSK, with 150 to 167 corrected errors. In terms of PAPR, PSK was more apt than QAM, but with a compromise in BER. The geometric losses were reduced with link length due to an increase in error correction capability for all three modulation cases, with the least losses occurring in 16-QAM. At the target bit error rate (BER), the signal to noise ratio (SNR) required for BPSK and QPSK was higher by 3.98 dB and 6.14 dB compared to 16-QAM
Study of Constrained Velocity Inversion of Seismic Data in North Sumatra Basin
Get PDFInterval velocity analysis increases the precision of seismic velocity data with a complex structure and a high variation of velocity both laterally and vertically. In this study, interval velocity analysis was performed by applying the exponential asymptotically bounded function approach. An exponentially asymptotically bounded function was applied to calculate the interval velocity obtained from the root mean square velocity of seismic data using the Dix equation for conversion. To control this operation, a velocity constraint was applied in the interval velocity conversion. The velocity constraint used was the velocity trend gained from the root mean square velocity. This method is called constrained velocity inversion. In this study, interval velocity analysis using constrained velocity inversion was applied to seismic data from the North Sumatra Basin area. The seismic data interpretation resulted from the interval velocity analysis using constrained velocity inversion described the subsurface structure clearly. A corresponding anomaly at a time depth from 2000 ms to 2400 ms in the seismic time-domain data indicated a fault beneath an anticline. This result indicates that the interval velocity analysis of seismic data is more rigorous than the root mean square velocity analysis