Penerapan Neural Network Dalam Klasifikasi Citra Permainan Batu Kertas Gunting dengan Probabilistic Neural Network

Siregar, Siti Julianita; Lubis, Ahmadi Irmansyah; Ginting, Erika Fahmi

doi:10.47065/bits.v3i3.1143

Cited by 3 publications

(4 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Beberapa penelitian terdahulu juga telah mengaplikasikan CNN dalam klasifikasi gambar dengan konteks yang mirip, seperti klasifikasi citra permainan batu-kertas-gunting (Naufal et al, 2021). dan klasifikasi citra permainan batu-kertas-gunting menggunakan Probabilistic Neural Network (Siregar et al, 2021). Namun, permasalahan yang umum terjadi dalam implementasi CNN adalah Overfitting terjadi ketika model terlalu cocok dengan data pelatihan dan kehilangan kemampuan untuk secara tepat mengeneralisasi terhadap data pengujian.. Pengembangan metode untuk mencegah overfitting menjadi hal yang sangat penting dalam meningkatkan performa model.…”

Section: Pendahuluanunclassified

“…Melihat penelitian-penelitian terdahulu dan tantangan yang dihadapi, penelitian ini bertujuan untuk mengembangkan metode klasifikasi gambar batu-kertas-gunting menggunakan Convolutional Neural Network (CNN) dengan penerapan fungsi callback untuk mencegah overfitting. Referensi dari penelitian terkait, seperti klasifikasi citra permainan batu-kertas-gunting (Fauzy et al, 2023) dan penggunaan CNN dalam konteks klasifikasi gambar (Siregar et al, 2021), akan menjadi landasan teoritis dalam pengembangan dan peningkatan metode yang diusulkan.…”

Section: Pendahuluanunclassified

See 1 more Smart Citation

Klasifikasi Gambar Batu-Kertas-Gunting Menggunakan Convolutional Neural Network dengan Fungsi Callback untuk Mencegah Overfitting

Samidin,

Fadjeri

2024

Jur. Penel. Inov.

View full text Add to dashboard Cite

Dalam era modern, kecerdasan buatan (AI) dan machine learning, khususnya teknik machine learning, telah menjadi integral dalam perkembangan teknologi. Implementasi Convolutional Neural Network (CNN) dalam klasifikasi gambar sering mengalami masalah overfitting, di mana model terlalu beradaptasi dengan data latih dan kurang mampu menggeneralisasi data uji. Untuk mencegah overfitting, penelitian ini menggunakan callback, seperti Early Stopping, Model Checkpointing, dan CSV Logging. Dataset gambar batu-kertas-gunting diperoleh dari Github Dicoding dan diproses dengan pembagian subset, augmentasi data, dan generator data. Model CNN dirancang dengan lapisan konvolusi, lapisan pooling, dan lapisan Dense. Optimizer, fungsi loss, dan metrik evaluasi dipilih sesuai tugas klasifikasi gambar. Melalui pelatihan model, akurasi meningkat dari 48.75% menjadi 98.75%, sementara loss menurun dari 0.9794 menjadi 0.0448. Evaluasi pada data validasi juga menunjukkan peningkatan konsisten. Pelatihan dihentikan pada epoch ke-18 setelah mencapai akurasi di atas 98%, untuk mencegah overfitting dan mengoptimalkan waktu pelatihan. Kesimpulan menunjukkan bahwa penggunaan callback efektif dalam mencegah overfitting dan meningkatkan akurasi model CNN dalam klasifikasi gambar batu-kertas-gunting. Ini menjadi strategi penting dalam meningkatkan kualitas hasil akhir model.

show abstract

Section: Pendahuluanunclassified

Klasifikasi Gambar Batu-Kertas-Gunting Menggunakan Convolutional Neural Network dengan Fungsi Callback untuk Mencegah Overfitting

Samidin,

Fadjeri

2024

Jur. Penel. Inov.

View full text Add to dashboard Cite

show abstract

“…The network paradigm also uses the Parzen estimations which adds up to builds density probabilities function required by Bayes theorem [26]. In addition, PNN has a structure consisting of four layers, namely the input layer and three information processing layers starting from the pattern layer to the summation layer, and continuing to the output layer [27].…”

Section: Modeling Classification Algorithmsmentioning

confidence: 99%

Comparison of Probabilistic Neural Network (PNN) and k-Nearest Neighbor (k-NN) Algorithms for Diabetes Classification

Azzahrah,

Alamsyah

2023

Recursive J. of Informatics

View full text Add to dashboard Cite

Purpose: This study aims to compare algorithms to determine the accuracy of the algorithm and determine the speed of the algorithm used for diabetes classification. Methods: There are two algorithms used in this study, namely Probabilistic Neural Network (PNN) and k-Nearest Neighbor (k-NN). The data used is the Pima Indians Diabetes Database. The data contains 768 data with 8 attributes and 1 target class, namely 0 for no diabetes and 1 for diabetes. The dataset has been divided into 80% training data and 20% testing data. Result: Accuracy is obtained after implementing k-fold cross validation with a value of k = 4. The accuracy results show that the k-Nearest Neighbor algorithm is superior and has better quickness compared to the Probabilistic Neural Network. The k-Nearest Neighbor algorithm obtains an accuracy of 74.6% for all features and 78.1% for four features Novelty: The novelty of this paper is optimizing and improving accuracy which is implemented with by focusing on data preprocessing, feature selection and k-fold cross validation in the classification algorithm

show abstract

“…This study uses three stages, namely lgrayscaling, canny edge detection and lthresholding. lGrayscale laims lto lmake lthe color uniformity lin lthe limage lso lthat lit lhas la lgray lvalue [10] , then lcanny ledge ldetection gives lresults lthat lmatch lthe image lline lpoints land lare lthe lresults lof lthe limage ledges [11] .…”

Section: Preprocessing and Segmentationmentioning

confidence: 99%

Classification of Alzheimer Disease from MRI Image Using Combination Naïve Bayes and Invariant Moment

Lubis¹,

Sibagariang²,

Ardi³

2023

Proceedings of the 5th International Conference on Applied Engineering, ICAE 2022, 5 October 2022, Batam, Indonesia

View full text Add to dashboard Cite

This study examines the classification of Alzheimer's disease. Alzheimer's is a memory disorder in an older person caused by degeneration of the central nervous system, which results in memory impairment and can cause death. Early detection of Alzheimer's can also be done based on image processing using magnetic resonance imaging (MRI) type images. Therefore, in this research, we use a feature extraction process to extract the characteristics of Alzheimer's disease that appear on MRI images using Moment Invariance and use the Naïve Bayes classification method to classify classes from images based on normal images, very mild disturbances, mild disturbances, and disturbances. Medium from brain images in the classification of Alzheimer's disease. The classification stage consists of several stages such as image acquisition, preprocessing (grayscaling), segmentation (canny edge detection, threshold), feature extraction using Invariant Moment, and image classification using Naïve Bayes. Based on the testing of the method proposed in this research, the results obtained for the accuracy of classifying Alzheimer's disease in this study are 94%.

show abstract

Penerapan Neural Network Dalam Klasifikasi Citra Permainan Batu Kertas Gunting dengan Probabilistic Neural Network

Cited by 3 publications

References 7 publications

Klasifikasi Gambar Batu-Kertas-Gunting Menggunakan Convolutional Neural Network dengan Fungsi Callback untuk Mencegah Overfitting

Klasifikasi Gambar Batu-Kertas-Gunting Menggunakan Convolutional Neural Network dengan Fungsi Callback untuk Mencegah Overfitting

Comparison of Probabilistic Neural Network (PNN) and k-Nearest Neighbor (k-NN) Algorithms for Diabetes Classification

Classification of Alzheimer Disease from MRI Image Using Combination Naïve Bayes and Invariant Moment

Contact Info

Product

Resources

About