Improving Convolutional Neural Network (CNN) Architecture (miniVGGNet) with Batch Normalization and Learning Rate Decay Factor for Image Classification

Ismail, A. S.; Ahmad, Siti Anom; Soh, Azura Che; Hassan, Khair; Harith, Hazreen Haizi

doi:10.30880/ijie.2019.11.04.006

Cited by 23 publications

(6 citation statements)

References 12 publications

(13 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For this study, we chose to use k-fold CV with 10 folds and repeated the process three times, for a total of 30 iterations of training and validation. With each repeat of the k-fold CV , the decay rate, a hyper-parameter used to minimize overfitting ( Ismail, Ahmad, Soh, Hassan, & Harith, 2019 ), was automatically manipulated within the caret package (0, .0001, and .1, respectively) to identify an optimum decay rate for model accuracy. As our decision was purely data-driven, based on the model performance, an optimum decay rate of .1 was selected for our model.…”

Section: New Methods In Combinationmentioning

confidence: 99%

“…For this study, we chose to use k-fold CV with 10 folds and repeated the process three times, for a total of 30 iterations of training and validation. With each repeat of the k-fold CV, the decay rate, a hyperparameter used to minimize overfitting (Ismail, Ahmad, Soh, Hassan, & Harith, 2019), was "The LDA technique is developed to transform the features into a lower dimensional space, which maximizes the ratio of the between-class variance to the within-class variance, thereby guaranteeing maximum separability" (Tharwat, Gaber, Ibrahim, & Hassanien, 2017, p. 170). Logistic Model Trees (LMT)…”

Section: Supervised Machine Learningmentioning

confidence: 99%

See 1 more Smart Citation

Using CATA and Machine Learning to Operationalize Old Constructs in New Ways: An Illustration Using U.S. Governors’ COVID-19 Press Briefings

Marshall

Yammarino

Parameswaran

et al. 2022

Organizational Research Methods

View full text Add to dashboard Cite

Increased computing power and greater access to online data have led to rapid growth in the use of computer-aided text analysis (CATA) and machine learning methods. Using “big data”, researchers have not only advanced new streams of research, but also new research methodologies. Noting this trend and simultaneously recognizing the value of traditional research methods, we lay out a methodology that bridges the gap between old and new approaches to operationalize old constructs in new ways. With a combination of web scraping, CATA, and supervised machine learning, using labeled ground truth data (i.e., data with known inputs and outputs), we train a model to predict CIP (Charismatic-Ideological-Pragmatic) leadership styles from running text. To illustrate this method, we apply the model to classify U.S. state governors’ COVID-19 press briefings according to their CIP leadership style. In addition, we demonstrate content and convergent validity of the method.

show abstract

Section: New Methods In Combinationmentioning

confidence: 99%

“…For this study, we chose to use k-fold CV with 10 folds and repeated the process three times, for a total of 30 iterations of training and validation. With each repeat of the k-fold CV, the decay rate, a hyperparameter used to minimize overfitting (Ismail, Ahmad, Soh, Hassan, & Harith, 2019), was "The LDA technique is developed to transform the features into a lower dimensional space, which maximizes the ratio of the between-class variance to the within-class variance, thereby guaranteeing maximum separability" (Tharwat, Gaber, Ibrahim, & Hassanien, 2017, p. 170). Logistic Model Trees (LMT)…”

Section: Supervised Machine Learningmentioning

confidence: 99%

Using CATA and Machine Learning to Operationalize Old Constructs in New Ways: An Illustration Using U.S. Governors’ COVID-19 Press Briefings

Marshall

Yammarino

Parameswaran

et al. 2022

Organizational Research Methods

View full text Add to dashboard Cite

show abstract

“…In the future, other biometrics characteristics such as iris and voice will be considered to be included in the system. Apart from that, Convolution Neural Network (CNN) will be also considered to be adopted as a classifier in order to improve the accuracies [25] [26]. ILDA is suggested to be improved to deal with other online learning problems.…”

Section: Discussionmentioning

confidence: 99%

Online Person Identification based on Multitask Learning

Anak Joseph,

Anak Pelias Pog,

Chin

et al. 2021

IJIE

View full text Add to dashboard Cite

In the digital world, everything is digitized and data are generated consecutively over the times. To deal with this situation, incremental learning plays an important role. One of the important applications that needs an incremental learning is person identification. On the other hand, password and code are no longer the only way to prevent the unauthorized person to access the information and it tends to be forgotten. Therefore, biometric characteristics system is introduced to solve the problems. However, recognition based on single biometric may not be effective, thus, multitask learning is needed. To solve the problems, incremental learning is applied for person identification based on multitask learning. Considering that the complete data is not possible to be collected at one time, online learning is adopted to update the system accordingly. Linear Discriminant Analysis (LDA) is used to create a feature space while Incremental LDA (ILDA) is adopted to update LDA. Through multitask learning, not only human faces are trained, but fingerprint images are trained in order to improve the performance. The performance of the system is evaluated by using 50 datasets which includes both male and female datasets. Experimental results demonstrate that the learning time of ILDA is faster than LDA. Apart from that, the learning accuracies are evaluated by using K-Nearest Neighbor (KNN) and achieve more than 80% for most of the simulation results. In the future, the system is suggested to be improved by using better sensor for all the biometrics. Other than that, incremental feature extraction is improved to deal with some other online learning problems.

show abstract

“…Specifically, the study limitation using this mobile application also being discussed and more plausibly these disliked factors are not determined by the physical factor rather than unexampled factor. Besides this proposed mobile augmented reality could be also be implemented using a neural network for future development and be able to classify the hand images by image classification [30].…”

Section: Discussionmentioning

confidence: 99%

Mobile Augmented Reality Hand Wash (MARHw): Mobile Application to guide community to ameliorate handwashing effectiveness to oppose Covid-19 disease

Hanafi¹,

Wahab²,

Wong³

et al. 2020

IJIE

View full text Add to dashboard Cite

The global concern about a new pandemic associated with the insufficiency of vaccines has necessitated basic practices of handwashing to get prevention in Covid-19 disease in modern health issues. At this point, due to that, the researchers develop a simple mobile application engaging augmented reality to educate the urban and rural community on a notable effect of basic infection prevention practices peculiarly on hand wash guidance to prevent Covid-19infection. A quality of a hand wash could be attained by washing hands in conformity with health standard guidelines. In this study, we analyze a sample size of 83 participants (n=83), there were categorized by age, young adult (ages 15 to 35, n=28), middle-age (36 to 55, n=28) and older adults (older than 56 years, n=27) through three main formulated hypotheses namely as a step, movement, and duration and being analyzed by SPSS version 25by using Anova. The differences in demographics and gender are not being compared amongst the group. The final result underlying a statistical significance draws two significant factors for step and movement and there was no significant effect at the time factor of utilizing this mobile application. Hence, the final result could mitigate the Malaysian government sector peculiarly to the health department to alleviate the load of educating a community about a proper hand wash that would ease the process.

show abstract

Improving Convolutional Neural Network (CNN) Architecture (miniVGGNet) with Batch Normalization and Learning Rate Decay Factor for Image Classification

Cited by 23 publications

References 12 publications

Using CATA and Machine Learning to Operationalize Old Constructs in New Ways: An Illustration Using U.S. Governors’ COVID-19 Press Briefings

Using CATA and Machine Learning to Operationalize Old Constructs in New Ways: An Illustration Using U.S. Governors’ COVID-19 Press Briefings

Online Person Identification based on Multitask Learning

Mobile Augmented Reality Hand Wash (MARHw): Mobile Application to guide community to ameliorate handwashing effectiveness to oppose Covid-19 disease

Contact Info

Product

Resources

About