Performance improvement of deep neural network classifiers by a simple training strategy

Çalışkan, Abdullah; Yuksel, MehmetEmin; Badem, Hasan; Baştürk, Alper

doi:10.1016/j.engappai.2017.09.002

Cited by 49 publications

(27 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Various studies have been done on the diabetes data classification using Pima Indian diabetes dataset. 3,5,8,9 Using the dataset from University of California, Irvine (UCI) machine learning repository, researchers used several methods for the classification problem and accuracy has been improved. A cascade learning system for diagnosis is proposed in Ref.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Type 2 diabetes data classification using stacked autoencoders in deep neural networks

Kannadasan

Edla

Kuppili

2019

Clinical Epidemiology and Global Health

118

View full text Add to dashboard Cite

This paper aims to classify the Pima Indians diabetes dataset with better accuracy and other evaluation metrics. The Deep Neural Network (DNN) framework will help to diagnose the patient in an effective way with higher accuracy. Method: In this approach, we proposed a Deep Neural Network framework for diabetes data classification using stacked autoencoders. Features are extracted from the dataset using stacked autoencoders and the dataset is classified using softmax layer. Also, fine tuning of the network is done using backpropagation in supervised fashion with the training dataset. However, the medical diagnosis involves the risk factors of wrong prediction; hence we have used evaluation metrics such as precision, recall, specificity and F1-score for the evaluation of our model and have achieved better results. Results: The proposed framework is experimented on Pima Indians Diabetes data which has 768 patient records with 8 attributes for each record. We achieved classification accuracy of 86.26%. Conclusion: A stacked autoencoders based Deep Learning framework for classification of Type 2 Diabetes data is proposed in this paper. This approach is experimented on UCI machine learning data and proved the outperformance over various existing classification methods.

show abstract

Section: Related Workmentioning

confidence: 99%

“…They achieved 77.09% classification accuracy for Pima Indian diabetes dataset. 9 A deep convolutional neural network based approach is proposed to classify the diabetic retinopathy cases with image dataset. 18 A prediction model is built using data mining techniques using improved k-means algorithm and logistic regression in Ref.…”

Section: Related Workmentioning

confidence: 99%

Type 2 diabetes data classification using stacked autoencoders in deep neural networks

Kannadasan

Edla

Kuppili

2019

Clinical Epidemiology and Global Health

118

View full text Add to dashboard Cite

show abstract

“…Bu farkın FOH'un bu iki yöntemden daha iyi olduğunu gösterip göstermediğine ancak istatistiksel analiz yapıldıktan sonra karar verilebilir. FOH, IOF ve MO üzerine kurgulanan bir yöntem olduğundan bu iki yöntemi istatistiksel olarak kıyaslamak için Wilcoxon işaretli sıra testi kullanılmıştır [23]. Tablo 4'de verilen sonuçlar incelendiğinde FOH lehine istatistiksel olarak anlamlı bir farklılık gözlenmektedir.…”

Section: IVunclassified

Emg Si̇nyalleri̇ İçi̇n Hi̇bri̇d Özni̇teli̇k Çikarma Yöntemi̇ Geli̇şti̇ri̇lmesi̇

Çalışkan¹

2019

Ömer Halisdemir Üniversitesi Mühendislik Bilimleri Dergisi

View full text Add to dashboard Cite

ÖZBu çalışmada EMG sinyalleri kullanılarak 14 farklı parmak hareketi önerilen yeni bir öznitelik çıkarma yöntemiyle sınıflandırılmıştır. EMG sinyallerinden parmak hareketlerinin tespiti / sınıflandırılması, ön işleme, özellik çıkarma ve sınıflandırma adımlarını içeren 3 ana adımdan oluşur. EMG sinyallerinin sınıflandırılmasında, sınıflandırıcının performansı doğrudan öznitelik çıkarma yöntemine bağlıdır. Bu sebeple öznitelik çıkarma yöntemi uygun bir biçimde seçilmelidir. Literatürde zaman, histogram ve frekans tabanlı birçok öznitelik çıkarma yöntemi vardır. Ancak bu yöntemler yüksek zaman karmaşıklığı, çok fazla işleme ihtiyaç duyma, fazla sayıda kontrol parametresi bulundurma, vb. birçok dezavantaja sahiptir. Bu çalışmada EMG sinyallerinden çeşitli parmak hareketlerinin sınıflandırılması amacıyla sayılan dezavantajları ortadan kaldırmak için yeni bir öznitelik çıkarma yöntemi sunulmuştur. Bu yöntem iki zaman tabanlı öznitelik çıkarma yönteminin hibritleştirilmesiyle elde edilmiştir. 10 kat çapraz doğrulamayla elde edilen deneysel sonuçlarına göre, önerilen yöntemin %97,48 doğruluk oranıyla bu çalışmada kullanılan diğer 9 öznitelik çıkarma yöntemlerinden daha iyi olduğu deneysel olarak gösterilmiştir. Deneysel sonuçlar istatistiksel yöntemlerle desteklenmiştir. ABSTRACTThis study aims to discuss classification of 14 different finger movements from EMG signals by using new feature extraction technique. The detection/classification of finger movements consists of 3 main steps including, preprocessing, feature extraction and classification steps. In classification of EMG signals, the performance of the classifier directly depends on feature extraction methods, including, time, histogram and frequency-based methods. However, these feature extraction methods have several drawbacks including, high time complexity, high computation demand, user supplied parameters, etc. In this paper, a new feature extraction method has been proposed for the classification of finger movements from EMG signals to overcome these problems. The proposed method based on hybridization of 2-time domain feature extraction techniques. The use of this method resulted in an accuracy of 97.48% after 10-fold-cross-validation. The experimental results supported with statistical analysis show that proposed method is better than 9 feature extraction methods investigated in this paper

show abstract

“…In medical database, techniques of data mining can be applied to the huge amount of historical data and derives precious rules for medical diagnosis [6]. The treatment, prevention and diagnosis can be aided through data mining [7][8][9][10]. Numerous data mining methods might be employed to cluster or classify the diseases.…”

Section: Introductionmentioning

confidence: 99%

Optimal Deep Learning based Data Classification Model for Type-2 Diabetes Mellitus Diagnosis and Prediction System

Ganesan*,

Sivakumar,

Thirumaran

et al. 2020

IJITEE

View full text Add to dashboard Cite

In recent days, deep learning models become a significant research area because of its applicability in diverse domains. In this paper, we employ an optimal deep neural network (DNN) based model for classifying diabetes disease. The DNN is employed for diagnosing the patient diseases effectively with better performance. To further improve the classifier efficiency, multilayer perceptron (MLP) is employed to remove the misclassified instance in the dataset. Then, the processed data is again provided as input to the DNN based classification model. The use of MLP significantly helps to remove the misclassified instances. The presented optimal data classification model is experimented on the PIMA Indians Diabetes dataset which holds the medical details of 768 patients under the presence of 8 attributes for every record. The obtained simulation results verified the superior nature of the presented model over the compared methods.

show abstract

Performance improvement of deep neural network classifiers by a simple training strategy

Cited by 49 publications

References 15 publications

Type 2 diabetes data classification using stacked autoencoders in deep neural networks

Type 2 diabetes data classification using stacked autoencoders in deep neural networks

Emg Si̇nyalleri̇ İçi̇n Hi̇bri̇d Özni̇teli̇k Çikarma Yöntemi̇ Geli̇şti̇ri̇lmesi̇

Optimal Deep Learning based Data Classification Model for Type-2 Diabetes Mellitus Diagnosis and Prediction System

Contact Info

Product

Resources

About