Percent Error Rate Research Articles

Prediction and Interpretability Study of the Glass Transition Temperature of Polyimide Based on Machine Learning with Quantitative Structure-Property Relationship (Tg-QSPR).

The glass transition temperature (Tg) is a crucial characteristic of polyimides (PIs). Developing a Tg predictive model using machine learning methodologies can facilitate the design of PI structures and expedite the development process. In this investigation, a data set comprising 1257 PIs was assembled, with Tg values determined using differential scanning calorimetry. 210 molecular descriptors were computed using RDKit, and subsequently, six distinct feature selection methodologies were employed to refine the descriptor set. Quantitative structure-property relationship models targeting Tg (Tg-QSPR) were then constructed using five ensemble learning algorithms and one deep learning algorithm. These models exhibited high predictive accuracy and robustness, with the CATBoost model demonstrating the highest accuracy, achieving a coefficient of determination of 0.823 for the test set, a mean absolute error of 20.1 °C, and a root-mean-square error of 29.0 °C. The study identified the NumRotatableBonds descriptor as particularly influential on Tg, showing a negative correlation with the property. Additionally, the model's accuracy was validated using ten new PI films not included in the original data set, resulting in absolute errors ranging from 2.5 to 26.9 °C and absolute percentage error rates of 1.0-12.8%. These findings underscore the importance of utilizing extensive and diverse data sets for predictive modeling to enhance accuracy and stability. Furthermore, exploring the interpretability of the model and experimentally validating newly synthesized PIs have augmented the practical utility of the model. Under the guidance of the Tg-QSPR models, it will be possible to accelerate the performance prediction and structural design of PIs in the future.

Modeling Multi-Output Back-Propagation DNN for Forecasting Indonesian Export-Import

Introduction/Main Objectives: International trade through the mechanisms of exports and imports plays a significant role in the Indonesian economy, making the timely availability of export and import value data crucial. Background Problems: Export and import values are influenced by inflation and exchange rate factors. Novelty: This study identifies two categories of variables, namely output (export value and import value) and input (inflation rate and the exchange rate of the Rupiah against the US Dollar). Research Methods: the research approach utilizes a Multi-output Deep Neural Network (DNN) with a Back-propagation algorithm to model the input-output relationship. The method can provide forecasting results for two or more bivariate or multivariate output variables. Finding/Results: The modeling analysis results indicate that the optimal model network structure is DNN (3.4). This model successfully predicts output 1 (export value) and output 2 (import value) with Mean Absolute Percentage Error (MAPE) rates of 13.76% and 13.63%, respectively. Additionally, the forecasting results show predicted export and import values for November to be US$ 16,208.13 billion and US$ 15,105.33 billion, respectively. These findings offer important insights into the direction of Indonesia's international trade movement, which can serve as a basis for future economic decision-making.

Design of adaptive protection coordination scheme using SVM for an AC microgrid

Microgrids have gathered a significant amount of attention within the past decade and becoming an essential asset in the energy industry. The ability to integrate sustainable renewable energy generation into the distribution network is one of the main reasons for microgrids' popularity. Besides technical advantages, renewable energy sources integration with the grid poses many technical challenges. Due to randomness in generation and different modes of operation, integrating renewable energy sources into power systems will be challenging. Reliable operation of renewable energy sources integrated microgrids even when faults occur, requires an adaptive protection scheme and relay coordination. In order to tackle those issues, this paper developed an adaptive protection coordination scheme using numerical overcurrent relays and support vector machines with a particle swarm optimization approach. Renewable energy sources based on microgrids use power electronics systems that require relays to operate in milliseconds, and the power system will be safe. particle swarm optimization's program execution time exceeds the primary relays. The machine learning algorithm is faster in computation time. This article proposes a method, particle swarm optimization, to estimate the optimum relay operating time. The machine learning model having a support vector machine is proposed to estimate the relay operating time. The proposed research is developed and tested on an IEEE12BUS and IEEE35BUS system. The support vector machine -based on adaptive protection coordination schemes estimates the operating time of directional overcurrent relays, with a lesser program execution time and a low percentage error rate. The derived result defined that the proposed approach is well-suitable for real-time applications of an AC microgrid system.

The accuracy of sonographic fetal weight in very preterm infants (≤32 weeks)

ObjectiveTo examine the accuracy of sonographic fetal weight to predict birthweight in very preterm infants (<32 weeks), and to compare the accuracy of estimated fetal weight (EFW) between those small for gestational age (SGA) and those appropriate for gestational age (AGA). Study designA retrospective study was conducted of data recorded between January 2010 and March 2023. Included were women with singleton livebirths at 23+0–31+6 weeks who had an EFW within one week from delivery. Mean percentage error, mean absolute percentage error, and underestimation and overestimation rates were calculated. We compared the accuracy of EFW between SGA and AGA infants. ResultsIn total, 360 women were included. The mean absolute percentage error was 7.8 % (range 0 %–68.9 %); for 207 (57.5 %) infants the percentage error was within ±10 %. Overestimation error >10 % was observed in 102 (28.3 %) infants and errors >20 % in 34 (9.4 %). Among infants born in the periviable period (23+0 – 25+6 weeks; N = 56), the mean absolute percentage error was 9.8 % (range: 0 %–40.3 %); the value was within ±10 % for only 28 periviable infants (50 %) and exceeded 20 % for 16.1 %. Among SGA compared to AGA infants, the mean absolute percentage error was higher (11.1% vs. 6.6 %, p = 0.035). Overestimation error >10 % was more frequent among SGA than AGA infants (55 (49.1 %) vs. 47 (19.0 %), p < 0.001). In a multivariate logistic regression analysis, SGA status was independently associated with a higher mean percentage error (beta = 0.260, p < 0.001) and an increased risk of an error >10 % (odds ratio = 2.1, 95 % confidence interval 1.2–3.5, p = 0.008). ConclusionsSonographic EFW is limited in assessing very preterm infants, particularly those who are SGA or born during the periviable period. These limitations should be considered regarding impending very preterm births and concerns about abnormal fetal growth.

Robust face recognition using quaternion interval type II fuzzy logic-based feature extraction on colour images.

In this paper, we propose a new robust and fast learning technique by investigating the effect of integration of quaternion and interval type II fuzzy logic along with non-iterative, parameter free deterministic learning machine (DLM) pertaining to face recognition problem. The traditional learning techniques did not account colour information and degree of pixel wise association of individual pixel of a colour face image in their network. Therefore, this paper presents a new technique named quaternion interval type II based deterministic learning machine (QIntTyII-DLM), which considers the interrelationship between three colour channels viz. red, green, and blue (RGB) by representing each colour pixel of a colour image in quaternion number sequence. Here, quaternion vector representation of a colour face image is fuzzified using interval type II fuzzy logic. This reduces the redundancy between pixels of different colour channels and also transforms colour channels of the image to orthogonal colour space. Thereafter, classification is performed using DLM. Experiments performed (on four standard datasets AR, Georgia Tech, Indian, face (female) and faces 94 (male) face datasets) and comparison done with other existing techniques proves that the proposed technique gives better results in terms of percentage error rate (reduces approximately 10-12%) and computational speed.

Development and Performance Evaluation of a Smart Upper-Limb Rehabilitation Exercise Device.

User-friendly rehabilitation medical devices can enhance health and the quality of life through the convergence of information communication and medical technology. Muscle contraction enables bodily movement, and the assessment of muscle strength is crucial. Muscle contraction includes isometric, isotonic, and isokinetic types. Many individuals with physical disabilities rely on wheelchairs due to lower-limb paralysis. There is a substantial correlation between the level of upper-limb functional recovery and the quality of daily life. This study aimed to design and evaluate a device that utilizes various muscle contractions to enhance the effectiveness of upper-limb rehabilitation exercises. The results from the isometric performance assessment showed percentage error rates of >30% for 5-30 kg. Correction equations were employed, and the isometric performance assessment resulted in error rates below 2.1% for 5-30 kg. Isokinetic performance assessment using video analysis evaluated a consistent wire speed from 10 cm/s to 70 cm/s with an average error rate of 0.91% across all speeds. This study demonstrates the ability to accurately measure various muscle contractions and showcases the potential for real-time feedback. This highlights how such a device could be helpful for target groups, including older adults and individuals with disabilities, during upper-limb rehabilitation exercises.

Development of chrono-spectral gold nanoparticle growth based plasmonic biosensor platform

Plasmonic sensor platforms are designed for rapid, label-free, and real-time detection and they excel as the next generation biosensors. However, current methods such as Surface Plasmon Resonance require expertise and well-equipped laboratory facilities. Simpler methods such as Localized Surface Plasmon Resonance (LSPR) overcome those limitations, though they lack sensitivity. Hence, sensitivity enhancement plays a crucial role in the future of plasmonic sensor platforms. Herein, a refractive index (RI) sensitivity enhancement methodology is reported utilizing growth of gold nanoparticles (GNPs) on solid support and it is backed up with artificial neural network (ANN) analysis. Sensor platform fabrication was initiated with GNP immobilization onto solid support; immobilized GNPs were then used as seeds for chrono-spectral growth, which was carried out using NH2OH at varied incubation times. The response to RI change of the platform was investigated with varied concentrations of sucrose and ethanol. The detection of bacteria E.coli BL21 was carried out for validation as a model microorganism and results showed that detection was possible at 102 CFU/ml. The data acquired by spectrophotometric measurements were analyzed by ANN and bacteria classification with percentage error rates near 0% was achieved. The proposed LSPR-based, label-free sensor application proved that the developed methodology promises utile sensitivity enhancement potential for similar sensor platforms.

A comparative evaluation of sliding contact tactile systems displaying braille and raised print

BACKGROUND: Access to low-cost tactile displays that allow sliding contact between text and reading fingers remains a challenge for blind and visually impaired (BVI) users. This impedes the widespread learning of braille and tactile reading. Previous work demonstrated a high accuracy in the tactile reading of braille and raised print presented at varying refresh rates. OBJECTIVE: This work compares the most suitable spacing between embossed characters on a sliding contact tactile display for the accurate reading of words. METHODS: Two discs, differing in inter-character spacing (ICS), embossed with braille on one side and raised print on the reverse side are used here. MNREAD sentences are read for a period of 5 minutes by 17 participants, who are visually impaired, using both discs. RESULTS: The results show that an ICS of 8 mm is sufficient for reading braille with a low percentage error rate of 6.4. However, an ICS of 8 mm does not allow similar rates while reading raised print. CONCLUSION: The results presented here will be relevant towards the research that works towards the design of economical sliding contact tactile displays for BVI users.

Spatio-Temporal Fusion of LiDAR and Camera Data for Omnidirectional Depth Perception

Object recognition and depth perception are two tightly coupled tasks that are indispensable for situational awareness. Most autonomous systems are able to perform these tasks by processing and integrating data streaming from a variety of sensors. The multiple hardware and sophisticated software architectures required to operate these systems makes them expensive to scale and operate. This paper implements a fast, monocular vision system that can be used for simultaneous object recognition and depth perception. We borrow from the architecture of a start-of-the-art object recognition system, YOLOv3, and extend its architecture by incorporating distances and modifying its loss functions and prediction vectors to enable it to multitask on both tasks. The vision system is trained on a large database acquired through the coupling of LiDAR measurements with complementary 360-degree camera to generate a high-fidelity labeled dataset. The performance of the multipurpose network is evaluated on a test dataset consisting of a total of 7,634 objects collected on a different road network. When compared with ground truth LiDAR data, the proposed network achieves a mean absolute percentage error rate of 11% on the passenger car within 10 m and a mean error rate of 7% or 9% on the truck within 10 m and beyond 10 m, respectively. It was also observed that adding a second task (depth perception) to the modeling network improved the accuracy of object detection by about 3%. The proposed multipurpose model can be used for the development of automated alert systems, traffic monitoring, and safety monitoring.

PENGARUH PEMAHAMAN WAJIB PAJAK, KUALITAS PELAYANAN PAJAK, DAN SANKSI PAJAK TERHADAP KEPATUHAN WAJIB PAJAK ORANG PRIBADI PADA KANTOR PELAYANAN PAJAK PRATAMA CIKARANG SELATAN

This study aims to examine and analyze the Effect of Taxpayer Understanding, Tax Service Quality and Tax Sanctions on Individual Taxpayer Compliance. The population in this study is individual taxpayers registered with KPP Pratama Cikarang Selatan. The determination of the number of samples in this study used the Slovin formula with a percentage error rate of 10%. with the number of samples used as many as 100 samples. The type of research applied to research is quantitative research through a descriptive approach. The data in this study uses primary data in the form of questionnaire distribution to individual taxpayers registered at KPP Pratama Cikarang Selatan who have income both employees / employees and entrepreneurs using Google forms that will be processed using the SPPS Version 25 program. The analytical methods used in this study are data quality test, classical assumption test, hypothesis test and multiple linear regression analysis. The results of this study show that Taxpayer Understanding, Tax Service Quality and Tax Sanctions simultaneously affect Individual Taxpayer Compliance. Viewed partially, Taxpayer Understanding and Tax Sanctions affect Individual Taxpayer Compliance. however, the Quality of Tax Services does not affect the Compliance of Individual Taxpayers. Thus, the better the understanding of taxpayers and tax sanctions will increase the compliance of individual taxpayers at the South Cikarang Pratama Tax Service Office, while the quality of tax services has not been able to fully improve the compliance of individual taxpayers

Longitudinal Study to Assess the Reliability and Predictive Validity of Ultrasound Measurements in Tracking Fetal Growth and Development Throughout Pregnancy

The utilization of ultrasound in obstetrics is extremely common due to its accurate fetal imaging capabilities. By early identifying anomalies such intrauterine growth restrictions and macrosomia, ultrasonography scanning during pregnancy primarily aims to reduce the risk of obstetric problems. Fetal weight is currently estimated using morphometric formulas. They employ fundamental biometric variables. Hadlock's formula, which is used for estimation of fetal weight, has 20 percent error rate, though. Due to this particular cause, the researchers from all around world had already searching for the additional sonographic characteristics having a stronger predictive value that correlate with fetal weight. According to recent scientific studies, assessing fetal weight can benefit from using novel parameters of sonograph such measurements of the thickness of soft tissue. The fetus's body can be measured in a number of locations, including upper arm, the thigh, abdomen, and the subscapular region. Diverse measurements have varying degrees of connection with some other anthropometric &amp; sonographic factors, including body mass &amp; gestational age. Numerous research using novel formulas for calculating fetal weight have been produced in response to the reports. For measuring fetal weight, measurements other than those including soft tissue, including such those lean and adipose tissue or utilising 3D ultrasonography are acquired. For the purpose of sonographic pregnancy assessment, ultrasound examination of the thicknesses of subcutaneous tissue in the several body regions might evidence to become reliable indicator of the fetal weight. Keywords: Pregnancy, postpartum hypoglycaemia, foetus, Intra-uterine, pregnancy ultrasound

Are Searches in OCR-generated Archives Trustworthy?

Abstract Digitised archives are revolutionary tools for research that, in a few seconds, generate results that earlier often took years to obtain. But do they provide all results for the terms searched for? The accuracy of searches was tested by performing sample searches of leading newspaper databases. The test revealed several weaknesses in the search process, including an average 18 percent error rate for single words in body text, and a far higher error rates for advertisements. Such high error rates encourage a critical look at the 20-year-old sector. Although these errors can be reduced by a re-digitation and with new improved OCR engines and new search algorithms, searches will nevertheless return manipulated results. In response, and to identify infringed bias and skewed representation, database owners need to provide thorough metadata to ensure source criticism.

DETECTING SUPPLIER PAYMENT ERRORS AND FRAUD USING A DATA WAREHOUSE AUDIT APPROACH

ABSTRACT This study provides management, accountants, and auditors with immediately applicable knowledge on supplier payment error discovery using a data warehouse. Despite the universality of supplier payment processing across government and non-government sectors, the associated error risks have received little attention in the academic literature and only limited attention in the practitioner literature. This paper seeks to articulate the data mining methods needed to perform such audits. A case study approach was used to analyze 15 client auditees in five countries using a variety of accounts payable systems. Action research methodology brought rigor to the development of incremental knowledge of error hypotheses and detection techniques. In this way, the prior knowledge gained from each case enriched the subsequent iterations. Even among clients whose management presumed human and system controls were robust, findings showed an average 0.19 percent error rate in payments to suppliers. These were mostly recoverable as cash refunds or credit notes. The audit cost-to-benefit ratio averaged 1:33, suggesting clients received financial returns 33 times higher than the cost of the audit, on average. Concurrently, internal controls were also hardened in accounts payable, tax compliance, and fraud prevention to secure additional future cost savings. Besides material cash recoveries and the occasional discovery of fraud, incremental lessons were also learned in data warehouse construction, data cleansing, privacy assurance, and tax compliance.

Analytical Framework of S-Parameter-Based Efficiency for Secondary-Parallel Compensation WPT System to Authenticate Data Using VNA

Resonant wireless power transfer (R-WPT) is the most efficient system in near-field applications. Predominantly, four distinct resonant topologies exist for R-WPT to enhance the power transfer efficiency (PTE) under different load conditions. Here, the PTE is considered crucial for gauging the R-WPT system performance. In a typical experimental scenario, measurement of the PTE is performed by exciting the transmitter (Tx) coil with a sinusoidal signal using a high-frequency inverter source. Therefore, the input and output power measurements are conducted using an oscilloscope. Here, the measurement using an oscilloscope is highly susceptible to external noise. As an alternative, the vector network analyzer (VNA) is the most accurate instrument to measure <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$S$ </tex-math></inline-formula> -parameters with high precision, even at very high frequencies. The <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$S$ </tex-math></inline-formula> -parameter-based analytical expressions for evaluating PTE are available only for series–series compensation topology. Therefore, VNA usage is severely limited for experimental validation of the other compensation topologies. This article proposes an analytical evaluation of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$S$ </tex-math></inline-formula> -parameter-based PTE for secondary-parallel compensation topology as a contribution. Moreover, the mathematical proof is deduced to determine the active operating load condition for the secondary-parallel capacitor. Error analysis of experimental data is performed where the signal-to-noise ratio (SNR) of the measurement system is greater than 35 dB and the percentage error rate <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\epsilon $ </tex-math></inline-formula> is less than 1.3%, which implies high precision and accuracy of the <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$S$ </tex-math></inline-formula> -parameter-based VNA measurement system.

Design of Security System Device for Motorized Vehicles through the Telegram Messenger Application and Updating GPS Locations on Smartphones in Real Time with IoT-based Smart Vehicles

Currently, the number of cases of motorized vehicle theft is increasing. Lack of double security, when left by the owner, is one of the factors that cause vehicles to be easily stolen. Based on these problems, we designed a motorized vehicle security system device that can monitor the condition and update the vehicle location with a smartphone through Global Positioning System (GPS). Furthermore, it can send an active alarm in the form of a buzzer alarm when there is an indication of the danger of vehicle theft. Furthermore, send notifications to the user so that the user immediately locks the vehicle by controlling it remotely in real-time through the Internet of Things (IoT)-based Smart Vehicle Security System using the Telegram Messenger Application and Google Maps (GMaps). The results of testing the response time show that the best performance is very responsive at 2.776 seconds in monitoring and controlling the vehicle. Moreover, the results of testing the vehicle distance position with GMaps and GPS obtained the best performance with a success rate percentage of 97.35% and an error rate percentage of 2.65%. It aims to make vehicle owners feel safe and comfortable and prevent motorized vehicle theft.

Classified 3D mapping and deep learning-aided signal power estimation architecture for the deployment of wireless communication systems

The traditional wireless communication systems deployment models require expensive and time-consuming procedures, including environment selection (rural, urban, and suburban), drive test data collection, and analysis of the raw data. These procedures mainly utilize stochastic and deterministic approaches for signal strength prediction to locate the optimum cellular tower (eNodeB) position for 4G and 5G systems. Since environment selection is limited by urban, suburban, and rural areas, they do not cover complex macro and micro variations, especially buildings and tree canopies having a higher impact on signal fading due to scattering and absorption. Therefore, they usually end up with high prediction errors. This article proposes an efficient architecture for the deployment of communication systems. The proposed method determines the effect of the environment via extracting tree and building properties by using a classified 3D map and You Only Look Once (YOLO) V5, which is one of the most efficient deep learning algorithms. According to the results, the mean average precision (mAP) 0.5% and mAP 0.95% accuracies are obtained as 0.96 and 0.45, and image color classification (ICC) findings indicate 77.6% accuracy on vegetation detection, especially for tree canopies. Thus, the obtained results significantly improved signal strength prediction with a 3.96% Mean Absolute Percentage Error (MAPE) rate, while other empirical models’ prediction errors fall in the range of 6.07–15.26%.

Concurrent Validity of the Garmin Vivofit®4 to Accurately Record Step Count in Older Adults in Challenging Environments.

There is little evidence of the concurrent validity of commercially available wrist-worn long battery life activity monitors to measure steps in older adults at slow speeds and with real-world challenges. Forty adults aged over 60years performed a treadmill protocol at four speeds, a 50-m indoor circuit, and a 200-m outdoor circuit with environmental challenges while wearing a Garmin Vivofit®4, the activPAL3™, and a chest-worn camera angled at the feet. The Garmin Vivofit®4 showed high intraclass correlation coefficients2,1 (.98-.99) and low absolute percentage error rates (<2%) at the fastest treadmill speeds and the outdoor circuit. Step counts were underestimated at the slowest treadmill speed and the indoor circuit. The Garmin Vivofit®4 is accurate for older adults at higher walking speeds and during outdoor walking. However, it underestimates steps at slow speeds and when walking indoors with postural transitions.

Asymmetrical Damage Aspects Based Investigations on the Disc Brake of Long-Range UAVs through Verified Computational Coupled Approaches

In recent years, the use of unmanned aerial vehicles (UAVs) has increased significantly. Asymmetrical factors, or frictional studies on the disc brake of UAVs, are one of the safety considerations taken into consideration during the design process because UAVs and their components have been built with the best safety in mind. This study focuses on choosing the optimal material for a UAV’s disc brake by using transient structural and thermal models. In order to compare the asymmetry-based frictional force produced by the two ways; the processes used in the transient simulation are validated using pin-on-disc (POD) testing. The foundation for this validation investigation is a metal matrix composite made of an aluminum alloy, and the basis tool is an ASTM G99-based computational test specimen. Steel-EN24 and carbon ceramic matrix composites testing are expanded using the same POD tests. A range of 3 percent to 8 percent error rates is found. As a result, the calculation techniques are applied to the UAV’s disc brake after they have proven to be trustworthy. This fixed-wing UAV’s extensions have a 5 kg payload capacity. The weight, avionics components, tire dimensions, and disc brake dimensions of the other UAV design parts are calculated using analytical formulas. The final designs are made using CATIA as a result. The grid convergence experiment is organized using a traditional finite element analysis tool. Finally, at its maximum rotational speed, a UAV’s disc brake is put through asymmetrical friction testing based on structural and thermal consequences. The correct materials for critical applications, such as carbon fiber-woven-wet-based reinforced polymer and Kevlar unidirectional-49-based reinforced polymer composites for changing rotating speeds, have now been made possible by fixed-wing UAVs.

Performance enhancement of visible light communication (VLC) system incorporating WMZCC‐OCDMA codes and PDM‐QPSK‐DSP data encoding

SummaryMulti‐user optical code division multiplexing (OCDMA) in visible light communication (VLC) is an imminent domain for providing an even distribution of available bandwidth and improved security. Q4R1Zero cross‐correlation (ZCC) codes are widely reported, and improved performance is offered by them due to insignificant multiple access interference (MAI). However, practically, MAI exists in ZCC codes due to adjacent weights (W) in the code matrix rows for each user. Therefore, a novel ZCC code with a nonadjacent W is designed in this work to eradicate MAI limitations and is named weight managed ZCC codes (WMZCC). An indoor 100 Gbps/User OCDMA‐VLC system is realised using polarisation division multiplexed quadrature phase shift keying (PDM‐QPSK) and digital signal processing (DSP). Incorporation of DSP in the coherent receiver compensates the effects of dispersion and nonlinear effects and also performs phase estimation. Further, a state‐of‐the‐art comparison of WMZCC codes with diagonal identity matrix (DIM) and zero cross‐correlation resultant weight (ZCCRW) codes has been performed in terms of log symbol error rate (SER), Q factor, error vector magnitude percentage (EVM%) and log bit error rate (BER) log at different VLC lengths, Q4R1Transmitter half angles (THA), incidence half angles (IHA) and optical concentrators (cm2). Constellations are analysed with and without DSP, and according to the author's best knowledge, a 100 Gbps VLC system for five users covering 6 m at −3.42 log BER is demonstrated for the first time. Results revealed that weight management in WMZCC codes makes its performance improve as compared with ZCCRW and DIM codes.

Acoustic Model with Multiple Lexicon Types for Indonesian Speech Recognition

Currently, speech recognition datasets are increasingly available freely in various languages. However, speech recognition datasets in the Indonesian language are still challenging to obtain. Consequently, research focusing on speech recognition is challenging to carry out. This research creates Indonesian speech recognition datasets from YouTube channels with subtitles by validating all utterances of downloaded audio to improve the data quality. The quality of the dataset was evaluated using a deep neural network. The time delay neural network (TDNN) was used to build the acoustic model by applying the alignment data from the Gaussian mixture model-hidden Markov model (GMM-HMM). Data augmentation was used to increase the number of validated datasets and enhance the performance of the acoustic model. The results show that the acoustic model built using the validated datasets is better than the unvalidated datasets for all types of lexicons. Utilizing the four lexicon types and increasing the data through augmentation to train the acoustic models can lower the word error rate percentage in the GMM-HMM, TDNN factorization (TDNNF), and CNN-TDNNF-augmented models to 40.85%, 24.96%, and 19.03%, respectively.