Abdulmajid Murad - Profile on Academia.edu (original) (raw)

Abdulmajid Murad

Daniel D. Hutto related author profile picture

Jonathan Zittrain related author profile picture

Bhubaneswar Mishra related author profile picture

Beat Signer related author profile picture

Alberto Pimpinelli related author profile picture

Alexandre Borovik related author profile picture

Jana  Javornik related author profile picture

Armando Marques-Guedes related author profile picture

Dan Zahavi related author profile picture

Graham Martin related author profile picture

Uploads

Papers by Abdulmajid Murad

Research paper thumbnail of Information-driven adaptive sensing based on deep reinforcement learning

Information-driven adaptive sensing based on deep reinforcement learning

Proceedings of the 10th International Conference on the Internet of Things

Research paper thumbnail of Deep Recurrent Neural Networks for Human Activity Recognition

Sensors

Adopting deep learning methods for human activity recognition has been effective in extracting di... more Adopting deep learning methods for human activity recognition has been effective in extracting discriminative features from raw input sequences acquired from body-worn sensors. Although human movements are encoded in a sequence of successive samples in time, typical machine learning methods perform recognition tasks without exploiting the temporal correlations between input data samples. Convolutional neural networks (CNNs) address this issue by using convolutions across a one-dimensional temporal sequence to capture dependencies among input data. However, the size of convolutional kernels restricts the captured range of dependencies between data samples. As a result, typical models are unadaptable to a wide range of activity-recognition configurations and require fixed-length input windows. In this paper, we propose the use of deep recurrent neural networks (DRNNs) for building recognition models that are capable of capturing long-range dependencies in variable-length input sequences. We present unidirectional, bidirectional, and cascaded architectures based on long short-term memory (LSTM) DRNNs and evaluate their effectiveness on miscellaneous benchmark datasets. Experimental results show that our proposed models outperform methods employing conventional machine learning, such as support vector machine (SVM) and k-nearest neighbors (KNN). Additionally, the proposed models yield better performance than other deep learning techniques, such as deep believe networks (DBNs) and CNNs.

Research paper thumbnail of DESIGN ND IMPLEMENTATION OF SOLAR TRACKING SYSTEM FOR PHOTOVOLTAIC CELLS

Currently, generating electricity by solar energy is inefficient and costly. The energy extracted... more Currently, generating electricity by solar energy is inefficient and costly. The energy extracted from a solar photovoltaic depends on solar insolation. For the extraction of maximum energy from the sun, the plane of a solar collector should always be normal to the incident radiation. In this graduation project, we seek to improve the solar system efficiency by designing and implementing an automatic solar tracking systems which will keep the solar panel aligned with the sun in order to maximize solar power extraction. This system tracks the maximum intensity of light by adjusting the solar panel to be normal to the incident light. This sun tracker system uses two light dependent resistors as sensors element to find the brightest point in the sky. The data and signal processing of light sensors are performed by a microcontroller based system which controls a rotating DC Motor. Performance of this system over the important parameters like solar radiation received on the collector, maximum hourly electrical power, efficiency gain, short circuit current, and open circuit voltage has been evaluated and compared with those for fixed tilt angle solar collector.

Research paper thumbnail of Information-driven adaptive sensing based on deep reinforcement learning

Information-driven adaptive sensing based on deep reinforcement learning

Proceedings of the 10th International Conference on the Internet of Things

Research paper thumbnail of Deep Recurrent Neural Networks for Human Activity Recognition

Sensors

Adopting deep learning methods for human activity recognition has been effective in extracting di... more Adopting deep learning methods for human activity recognition has been effective in extracting discriminative features from raw input sequences acquired from body-worn sensors. Although human movements are encoded in a sequence of successive samples in time, typical machine learning methods perform recognition tasks without exploiting the temporal correlations between input data samples. Convolutional neural networks (CNNs) address this issue by using convolutions across a one-dimensional temporal sequence to capture dependencies among input data. However, the size of convolutional kernels restricts the captured range of dependencies between data samples. As a result, typical models are unadaptable to a wide range of activity-recognition configurations and require fixed-length input windows. In this paper, we propose the use of deep recurrent neural networks (DRNNs) for building recognition models that are capable of capturing long-range dependencies in variable-length input sequences. We present unidirectional, bidirectional, and cascaded architectures based on long short-term memory (LSTM) DRNNs and evaluate their effectiveness on miscellaneous benchmark datasets. Experimental results show that our proposed models outperform methods employing conventional machine learning, such as support vector machine (SVM) and k-nearest neighbors (KNN). Additionally, the proposed models yield better performance than other deep learning techniques, such as deep believe networks (DBNs) and CNNs.

Research paper thumbnail of DESIGN ND IMPLEMENTATION OF SOLAR TRACKING SYSTEM FOR PHOTOVOLTAIC CELLS

Currently, generating electricity by solar energy is inefficient and costly. The energy extracted... more Currently, generating electricity by solar energy is inefficient and costly. The energy extracted from a solar photovoltaic depends on solar insolation. For the extraction of maximum energy from the sun, the plane of a solar collector should always be normal to the incident radiation. In this graduation project, we seek to improve the solar system efficiency by designing and implementing an automatic solar tracking systems which will keep the solar panel aligned with the sun in order to maximize solar power extraction. This system tracks the maximum intensity of light by adjusting the solar panel to be normal to the incident light. This sun tracker system uses two light dependent resistors as sensors element to find the brightest point in the sky. The data and signal processing of light sensors are performed by a microcontroller based system which controls a rotating DC Motor. Performance of this system over the important parameters like solar radiation received on the collector, maximum hourly electrical power, efficiency gain, short circuit current, and open circuit voltage has been evaluated and compared with those for fixed tilt angle solar collector.

Log In