Effectiveness of Backpropagation Algorithm in Healthcare Data Classification
Nowadays, researchers are trying to reveal better consequences by acting on machine learning (ML) algorithms. The notion behind this study is to represent the fundamental machine learning algorithms and its applicability in current scenario. Backpropagati
- PDF / 233,874 Bytes
- 10 Pages / 439.37 x 666.142 pts Page_size
- 4 Downloads / 166 Views
Abstract Nowadays, researchers are trying to reveal better consequences by acting on machine learning (ML) algorithms. The notion behind this study is to represent the fundamental machine learning algorithms and its applicability in current scenario. Backpropagation is considered as one of the classic supervised algorithms for training and classifying the feedforward neural networks. The concept of backpropagation used as a means in neural networks for transmitting entire error back to lessen the loss is termed as backpropagation network (BPN). We have considered BPN for classification as it is flexible, less complex, and performs better with noise-free data. The experimental analysis has been carried out by gathering dataset from UCI storehouse. Popular datasets like cancer, diabetes, heart, and liver are chosen for study. The classifier efficiency has been shown by observing its lower RMSE value and better accuracy with other factors also. By developing a BPN-based classifier system, it may ascertain physicians to deal with health-related problems. Keywords Machine learning · Neural network · Backpropagation neural network · Classification · Optimization
C. Chandra Sekhar · N. Panda (B) · B. V. Ramana · B. Maneesha · S. Vandana Department of Information Technology, Aditya Institute of Technology and Management (AITAM), Tekkali, K Kotturu 532201, Andhra Pradesh, India e-mail: [email protected] C. Chandra Sekhar e-mail: [email protected] B. V. Ramana e-mail: [email protected] B. Maneesha e-mail: [email protected] S. Vandana e-mail: [email protected] © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 R. Sharma et al. (eds.), Green Technology for Smart City and Society, Lecture Notes in Networks and Systems 151, https://doi.org/10.1007/978-981-15-8218-9_25
289
290
C. Chandra Sekhar et al.
1 Introduction Machine learning (ML) is an application of artificial intelligence which has flexibility to automatically acquire and progress from competence while not being specifically programmed [1]. The main concentration is on schedule of computer programs which access available information to learn for itself, and the same concept is associated with computational measurements and numerical optimization [2]. Multiple methods of ML are predominantly used across disciplines, such as supervised learning, unsupervised learning, semi-supervised learning, and reinforcement learning. These are used to automate various tasks which are considered to be complex jobs for humans such as recognition of image and generating texts and different Web-based games [3]. ML is going to have a huge impact on the economy and on life in general. Some of the use cases are manufacturing, retail, energy demand, and supply chain optimization. It also uses the information to train an appliance to generate patterns, and depending on nature and type of pattern, the machines can able to take further precise resolutions while dealing with new-fangled data [4]. Clas
Data Loading...
