Enhanced Classification of Motor Imaginary in EEG Using Feature Optimization and Machine Learning | Chapter 3 | Science and Technology: Developments and Applications Vol. 6

Accurate classification of motor imagery (MI) in EEG signals plays a crucial role in the diagnosis of neurological diseases, including conditions affecting motor control such as brain strokes and amyotrophic lateral sclerosis (ALS). However, the complex and high-dimensional nature of MI-EEG data poses significant challenges for accurate classification. Traditional classification methods often struggle with noise, artifacts, and redundant features, leading to reduced classification accuracy and increased computational complexity.

This paper presents an enhanced classification technique leveraging feature optimization and a deep neural network (DNN) classifier to improve the accuracy of MI-EEG data classification. The proposed approach utilizes a three-layer DNN model integrated with the Teacher Learning-Based Optimization (TLBO) technique. This optimization method reduces noise and artifacts in EEG signals, enhancing the quality of input vectors for the DNN classifier. The feature extraction process employs discrete wavelet transform (DWT) to decompose the EEG signals into multiple sub-bands, capturing essential frequency components. Subsequently, the TLBO algorithm refines these features, optimizing them for improved classification performance.

The proposed algorithm was evaluated using datasets from the third and fourth BCI competitions and simulated within a MATLAB environment. Comparative analysis was conducted against existing algorithms, including Bayesian Networks (BN) and Ensembled Machine Learning (EBL), to validate the performance of the proposed method. Experimental results demonstrate that the suggested approach significantly improves classification accuracy across various EEG signal bands, including raw, delta, theta, alpha, and beta signals. The combination of DNN with TLBO not only enhances classification accuracy but also reduces computational complexity by selecting the most relevant features.

The findings highlight the potential of the proposed approach in developing robust and reliable MI-based Brain-Computer Interface (BCI) applications for motor control, such as assistive communication systems, gaming, and wheelchair control for individuals with motor disabilities. Future work will focus on extending this approach to classify multi-class MI tasks, thereby broadening its applicability in advanced communication and control systems.

 

Author (s) Details

 

Virendra Kumar Tiwari
Department of Computer Application, Lakshmi Narain College of Technology (MCA), Bhopal, MP-462022, India.

 

Priyanka Singha
Department of Computer Science Engineering, Lakshmi Narain College of Technology Excellence, Bhopal, India.

 

Sonal Sharma
Department of Computer Application, Lakshmi Narain College of Technology (MCA), Bhopal, MP-462022, India.

 

Anshu Gangwar
Department of Computer Application, Lakshmi Narain College of Technology (MCA), Bhopal, MP-462022, India.

 

Please see the book here:- https://doi.org/10.9734/bpi/stda/v6/4324

Method for Calculating 3D Coordinates in Pose Estimation of a Stick Held by a Hand | Chapter 4 | Advances and Challenges in Science and Technology Vol. 4

 Monocular human 3D pose belief has become more proficient with deep interconnected system (DNN) technology. One of the next challenges is pose belief including stick grasped in the help. In this study, we propose a order to compute the matches of the stick tip using a human 3D pose belief network that measures the position where the stick is grasped in the hand. In particular, when the stick is four-sided to the camera's line-of-sight to the tip point, the relates of the measurement point concede possibility be imaginary, and we present a resolution using the system of Lagrange multiplier. Experiments habitual the ability of the projected method to reduce the solution to a honest number when it would result in an invented solution. The projected method be necessary to be used to monocular motion capture of humans assets differing stick-shaped objects, e.g., in sword fighting methods and baseball batting.

Author(s) Details:

Kazumoto Tanaka,
Faculty of Engineering, Kindai University, Higashi-Hiroshima, Japan.

Please see the link here: https://stm.bookpi.org/ACST-V4/article/view/12096

A Simple Root Location Estimation for 3D Pose Estimation with Root-Relative Coordinate System | Chapter 8 | Research and Developments in Engineering Research Vol. 8

 There is a powerful need for human action acknowledgment in various areas of organization. One of the most main technologies for physical operation recognition is human pose following. In recent years, three-spatial (3D) pose tracking means based on deep affecting animate nerve organs networks (DNNs) have made significant progress. However, most of these plans require a expensive computer equipped accompanying a graphics convert unit (GPU). The purpose of this study search out provide a inconsequential pose estimation network for human operation recognition tasks. We used Google's MediaPipe Pose, that is optimized to arrest real time on the CPU alone, for operation recognition on a depressed-end personal computer (PC). However, MediaPipe Pose cannot track the neighborhood of a human body cause it estimates poses on a coordinate system accompanying the waist as the origin (that is, a root-relative coordinate system). Therefore, in this place study, we developed a method to get the absolute matches of the root with a simple estimate. It was also establish that moving median permeating on the computed 3D absolute relates of the root can reduce shake in pose tracking. The proposed order can be acted in near real-opportunity on a commercially available PC accompanying a camera, and is expected to have many requests.

Author(s) Details:

Kazumoto Tanaka,
Faculty of Engineering, Kindai University, Higashi-Hiroshima, Japan.

Please see the link here: https://stm.bookpi.org/RADER-V8/article/view/11956

Efficient Synergetic Filtering in Big Dataset Using Deep Neural Network Technique | Chapter 12 | Advanced Aspects of Engineering Research Vol. 11

 Speech recognition, computer vision, and natural language processing have all benefited from deep neural networks. In this mission, we concentrated on neural network techniques to address the major challenge in synergetic or collaborative -filtering based on the concept of hidden feedback. While deep learning has been employed in a few recent research, it has primarily been employed to create supplementary facts like textual metaphors for things and the acoustic capabilities of music. Matrix factorization is still used for the most significant part of synergetic filtering, communication between customer and object capabilities, and a core product based on hidden customer and object capabilities has been introduced. Artificial Neural Synergetic Filtering (ANSF) is a typical framework for replacing the fundamental makeup with a neural design that can be very efficient in analysing data using a random function. The ANSF is a prominent matrix-factorization framework that is both common and potentially unique. To improve ANSF modelling with non-linearities, we propose utilising a multi-layer perceptron to examine the customer–object contact mechanism. Experiments on real worldwide databases reveal that our suggested ANSF outperforms current approaches significantly. According to research findings, using core layers of artificial neural networks improves overall efficiency. This work improves on existing shallow models for synergetic filtering, paving the way for a new line of research into deep learning-based recommendation.

Author (s) Details

B. Mukunthan
Department of Computer Science, Sri Ramakrishna College of Arts and Science (Autonomous), India, Tamil Nadu, Coimbatore-641021, India.

M. Arunkrishna
PG & Research Department of Computer Science, Jairams Arts and Science College (Affiliated to Bharathidhasan University), Karur - 639003, Tamilnadu, India.

View Book : https://stm.bookpi.org/AAER-V11/article/view/1257

Recent Study on Breast Cancer Prediction Based on Deep Neural Network Model Implemented AWS Machine Learning Platform | Chapter 3 | Advanced Aspects of Engineering Research Vol. 11

 Breast cancer is one of the most severe tumours in women, and developing breast tissue can result in mortality. Surgery, radiation, chemicals in combination with hormone therapy, and biological therapy are only a few of the current treatments for breast cancer that have shown to be effective. This paper compares the Deep Neural Network (DNN) model to other machine learning approaches such as XGBoost and Random Forest on a public dataset using the AWS machine learning framework. The plot of model accuracy for the training and validation sets, as well as performance assessment metrics to evaluate the model, reveal that the DNN model with Hyperparameter tweaking produces the best results for breast cancer prediction.

Author (s) Details

Le Dinh Phu Cuong
College of Computer Science and Electronic Engineering, Hunan University, Changsha 410082, China and Yersin University, Vietnam.

Dong Wang
College of Computer Science and Electronic Engineering, Hunan University, Changsha 410082, China.

Duyen The Hoang
College of Computer Science and Electronic Engineering, Hunan University, Changsha 410082, China.

Le Mai Nhu Uyen
Yersin University, Vietnam and College of Life Science, Hunan Normal University, Changsha 410082, China.

View Book : https://stm.bookpi.org/AAER-V11/article/view/1244