Ammar Falih Mahdi
@ruc.edu.iq
Computer Science / E-Learning Unit
AL - Rafidain University College
EDUCATION
PHD in Computer Science - Artificial Intelligence and Web Programming, M.Sc. in Computer Science - Artificial Intelligence, B.Sc. in Computer Science.
RESEARCH INTERESTS
Artificial Intelligence, Web Application Programming, Mobile Programming
Scopus Publications
Scholar Citations
Scholar h-index
Scopus Publications
Mustafa Mohammmed Jassim, Haithem Kareem Abass, Azhar Raheem Mohammed Al-Ani, Ammar Falih Mahdi, Abdul Mohsen Jaber Almaaly, Yuliia Navrozova, and Nataliia Bodnar
IEEE
Background: Climate change stands as one of the most critical global challenges with enormous implications. Since its science is well understood, efforts now focus on modeling and predicting its effects to mitigate or adapt to these. Deep learning, with its remarkable aptitude for data representation and analysis, is a promising candidate to enhance weather attack predictions on a global scale.Objective: This empirical study will assess their central tendencies and relationships for understanding the effectiveness of deep learning models in anticipation of climate change impacts. The paper investigates whether recently proposed models can provide better predictions than traditional techniques.Methodology: Authors utilize a detailed dataset of past climate data and its consequences This dataset is used to train and test deep learning architectures, such as Convolutional Neural Networks (CNN) or Recurrent Neural Networks (RNN), but, for the first, comparing with traditional regression models.Results: The findings show that DL techniques are very effective in comparison to traditional methods when it comes to predicting the impacts of climate change. Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs) have proven to be highly accurate at detecting complex relationships among climate factors and their impacts such as extreme weather events or sea level rise.Conclusion: The potential of deep learning approaches to improve our ability to model the consequences of climate change is substantial. Its forecasts have greater skill and the ability to inform policy and adaptation effectively. Given the continued acceleration of climate change, deploying advanced machine learning will be critical to maintaining a steady state.
Ammar Falih Mahdi and Aseel Khalid Ahmed
Institute of Advanced Engineering and Science
One of the most common causes of functional frailty is major depressive disorder (MDD). MDD is a chronic condition that requires long-term therapy and professional assistance. Additionally, MDD effective treatment requires early detection. Unfortunately, it has intricated clinical characteristics that make early diagnosis and treatment difficult for clinicians. Furthermore, there are currently no clinically effective diagnostic biomarkers that can confirm an MDD diagnosis. However, electroencephalogram (EEG) data from the brain have recently been used to make a quantitative diagnosis of MDD. In addition, As being among the most cutting-edge artificial intelligence (AI) technologies, deep learning (DL) has exhibited superior performance in a wide range of real-world applications, from computer vision to healthcare. However, an additional challenge could be the extraction of information from the ECG raw data. This paper presents a method for converting EEG data to power spectral density (PSD) images, and then they were classified as healthy or MDD using a deep neural network for feature extraction and a machine learning (ML) classifier. When employing the proposed approach, the images formed from the PSD show a considerably improved performance in classification results.