Simulation of infant suction vacuum pressure using WGAN

Fatih Furkan Arslan; Önder Dinçel

Makale detayları

Simulation of infant suction vacuum pressure using WGAN

Yazarlar:

Fatih Furkan Arslan | Önder Dinçel

Yayıncı:

Yozgat Bozok University

Ana Başlıklar:

Derin Öğrenme

Dergi:

Bozok Journal of Engineering and Architecture Vol.3 Issue 2, pp.78-86

Belge Tipi:

Makale

Yayın Tarihi:

12/2024

Belge Dili:

İngilizce

Document ID (Org No.):

2001331176

pISSN:

3023-4298

Anahtar kelimeler:

Özet:

 In this study, it is aimed to simulate infant suction vacuums by replicating infant suction vacuum pressures with the Wasserstein Generative Adversarial Network WGAN model. To understand the physiology of infants feeding behavior, it is very important to understand the physiology of sucking. One of the most important elements of feeding physiology is the infant suction vacuum. In this context, the infant suction vacuum pressure produced by the WGAN model was successfully simulated. Data preprocessing and hyperparameter optimization were performed to ensure the most accurate simulation of the WGAN model used in the study. As a result, the low RMSE and high R2 value between the real data and the generated data show that the model works successfully. This method is promising in terms of better understanding the sucking behavior of infants and contributing to solving problems in the field of infant feeding. In the future, infant sucking simulation will be more successful if the model is supported with larger data and other elements of infant sucking physiology such as swallowing and respiration are considered. This kind of simulation will pave the way for new applications in the fields of infant feeding and breastfeeding.

 In this study, it is aimed to simulate infant suction vacuums by replicating infant suction vacuum pressures with the Wasserstein Generative Adversarial Network WGAN model. To understand the physiology of infants feeding behavior, it is very important to understand the physiology of sucking. One of the most important elements of feeding physiology is the infant suction vacuum. In this context, the infant suction vacuum pressure produced by the WGAN model was successfully simulated. Data preprocessing and hyperparameter optimization were performed to ensure the most accurate simulation of the WGAN model used in the study. As a result, the low RMSE and high R2 value between the real data and the generated data show that the model works successfully. This method is promising in terms of better understanding the sucking behavior of infants and contributing to solving problems in the field of infant feeding. In the future, infant sucking simulation will be more successful if the model is supported with larger data and other elements of infant sucking physiology such as swallowing and respiration are considered. This kind of simulation will pave the way for new applications in the fields of infant feeding and breastfeeding.

 In this study, it is aimed to simulate infant suction vacuums by replicating infant suction vacuum pressures with the Wasserstein Generative Adversarial Network WGAN model. To understand the physiology of infants feeding behavior, it is very important to understand the physiology of sucking. One of the most important elements of feeding physiology is the infant suction vacuum. In this context, the infant suction vacuum pressure produced by the WGAN model was successfully simulated. Data preprocessing and hyperparameter optimization were performed to ensure the most accurate simulation of the WGAN model used in the study. As a result, the low RMSE and high R2 value between the real data and the generated data show that the model works successfully. This method is promising in terms of better understanding the sucking behavior of infants and contributing to solving problems in the field of infant feeding. In the future, infant sucking simulation will be more successful if the model is supported with larger data and other elements of infant sucking physiology such as swallowing and respiration are considered. This kind of simulation will pave the way for new applications in the fields of infant feeding and breastfeeding.

 Bu çalışmada, bebeklerin emme vakum basınçlarının Wasserstein Generative Adversarial Network WGAN modeli ile çoğaltılarak bebek emme vakumlarının benzetiminin yapılması amaçlanmıştır. Bebeklerin beslenme davranışlarını fizyolojisini anlamak için emme fizyolojisini kavramak oldukça önemlidir. Beslenme fizyolojisinin en önemli unsurlarından bir tanesi bebek emme vakumudur. Bu bağlamda, WGAN modelinin ile üretilen bebek emme vakum basıncı başarıyla taklit edilmiştir. Çalışmada kullanılan WGAN modelinin en doğru benzetimi yapabilmesi için veri ön işleme yapılmış ve hiperparametre optimizasyonu yapılmıştır. Sonuç olarak, gerçek verilerle üretilen veriler arasındaki düşük RMSE ve yüksek R2 değeri modelin başarılı çalıştığını göstermektedir. Bu yöntem ile bebeklerin emme davranışlarının daha iyi anlaşılması ve bebek beslenmesi alanındaki sorunların çözülmesine katkı sağlaması açısından umut vadetmektedir. Gelecekte, modelin daha geniş verilerle desteklenmesi ve bebek emme fizyolojisinin diğer unsurları olan yutma ve solunum gibi unsurlarında ele alınmasıyla bebek emme benzetim daha başarılı bir şekilde yapılabilecektir. Bu tür bir benzetim ile bebek beslenmesi ve anne sütü sağımı alanlarında yeni uygulamalara zemin hazırlayacaktır.

Makale detayları

Simulation of infant suction vacuum pressure using WGAN

Yazarlar:

Fatih Furkan Arslan | Önder Dinçel

Yayıncı:

Yozgat Bozok University

Ana Başlıklar:

Derin Öğrenme

Dergi:

Bozok Journal of Engineering and Architecture Vol.3 Issue 2, pp.78-86

Belge Tipi:

Makale

Yayın Tarihi:

12/2024

Belge Dili:

İngilizce

Document ID (Org No.):

2001331176

pISSN:

3023-4298

Infant Sucking Vacuum

WGAN

Simulation

Sucking PhysiologyBebek Emme Vakumu

WGAN

Benzetim

Emme Fizyolojisi

Sayı

Makale detayları

Yazarlar: Fatih Furkan Arslan | Önder Dinçel

Yayıncı: Yozgat Bozok University

Ana Başlıklar: Derin Öğrenme

Dergi: Bozok Journal of Engineering and Architecture Vol.3 Issue 2, pp.78-86

Belge Tipi: Makale

Yayın Tarihi: 12/2024

Belge Dili: İngilizce

Document ID (Org No.): 2001331176

pISSN: 3023-4298

Sayı

Yazarlar:

Fatih Furkan Arslan | Önder Dinçel

Yayıncı:

Yozgat Bozok University

Ana Başlıklar:

Derin Öğrenme

Dergi:

Bozok Journal of Engineering and Architecture Vol.3 Issue 2, pp.78-86

Belge Tipi:

Makale

Yayın Tarihi:

12/2024

Belge Dili:

İngilizce

Document ID (Org No.):

2001331176

pISSN:

3023-4298