231 Hits in 1.5 sec

Deep Belief Networks for Image Denoising [article]

Mohammad Ali Keyvanrad, Mohammad Pezeshki, Mohammad Ali Homayounpour
2014 arXiv   pre-print
Deep Belief Networks which are hierarchical generative models are effective tools for feature representation and extraction. Furthermore, DBNs can be used in numerous aspects of Machine Learning such as image denoising. In this paper, we propose a novel method for image denoising which relies on the DBNs' ability in feature representation. This work is based upon learning of the noise behavior. Generally, features which are extracted using DBNs are presented as the values of the last layer
more » ... . We train a DBN a way that the network totally distinguishes between nodes presenting noise and nodes presenting image content in the last later of DBN, i.e. the nodes in the last layer of trained DBN are divided into two distinct groups of nodes. After detecting the nodes which are presenting the noise, we are able to make the noise nodes inactive and reconstruct a noiseless image. In section 4 we explore the results of applying this method on the MNIST dataset of handwritten digits which is corrupted with additive white Gaussian noise (AWGN). A reduction of 65.9% in average mean square error (MSE) was achieved when the proposed method was used for the reconstruction of the noisy images.
arXiv:1312.6158v2 fatcat:736mol5twjdrloa3gc23lvzy2i

Sequence Modeling using Gated Recurrent Neural Networks [article]

Mohammad Pezeshki
2015 arXiv   pre-print
In this paper, we have used Recurrent Neural Networks to capture and model human motion data and generate motions by prediction of the next immediate data point at each time-step. Our RNN is armed with recently proposed Gated Recurrent Units which has shown promising results in some sequence modeling problems such as Machine Translation and Speech Synthesis. We demonstrate that this model is able to capture long-term dependencies in data and generate realistic motions.
arXiv:1501.00299v1 fatcat:tt2w4lymovhezbcl7hdpwxgbta

Unnecessary Overtreatment Index: a useful tool in family practice

Mohammad Zakaria Pezeshki, Sina Pezeshki
2015 Revista Brasileira de Medicina de Família e Comunidade  
Cite as: Pezeshki MZ, Pezeshki S. Unnecessary Overtreatment Index: a useful tool in family practice. Rev Bras Med Fam Comunidade. 2015;10(35):1-2.  ... 
doi:10.5712/rbmfc10(35)1066 fatcat:zah5dm7ierdwtgbm6ayxuwr6eu

ٍEducating Quaternary Prevention (P4) in Iran for decreasing the harms and costs of unnecessary services in clinical medicine and public health

Mohammad Zakaria Pezeshki, Sina Pezeshki
2013 Payesh  
In Clinical Medicine and Public Health, Primary, Secondary and Tertiary Preventive services mainly focus on preventing of diseases. But Quaternary Prevention (P4) is related to preventing of unnecessary diagnostic, therapeutic and preventative services. P4 was coined by Marc Jamoulle, a Belgian family physician, in 1986. Unfortunately P4 is not well known in Iran. In this review the history, definition and goals of P4 is presented. As some of the unnecessary clinical medicine and public health
more » ... ervices are prevalent in Iran, the importance of educating P4 in Iranian universities of medical sciences and health care services is discussed.
doaj:081a8b088b9b4bdc883f627103aa939b fatcat:63yomxjy2jaibbtisjiuslg6de

Multi-scale Feature Learning Dynamics: Insights for Double Descent [article]

Mohammad Pezeshki, Amartya Mitra, Yoshua Bengio, Guillaume Lajoie
2021 arXiv   pre-print
Mohammad Pezeshki, Sékou-Oumar Kaba, Yoshua Bengio, Aaron Courville, Doina Precup, and Guillaume Lajoie.  ...  Mohammad Pezeshki, Sékou-Oumar Kaba, Yoshua Bengio, Aaron C. Courville, Doina Precup, and Guillaume Lajoie. Gradient starvation: A learning proclivity in neural networks.  ... 
arXiv:2112.03215v1 fatcat:7kshshaudrdizblhuyb3btf6a4

consumer segmentation on the basis of food-related lifestyles

Sayyed Mohammad Tabatabaee Nasab, Fereshteh Pezeshki Najafabadi
2016 ‫مدیریت بازرگانی  
Effective Competition in today's markets requires correct identification of market segments. The food market is one of markets that need to be segmented on the basis of appropriate criteria. Food-Related Life Styles (FRL) is a comprehensive tool for segmentation of the food market in Europe that has rarely been used in Iran. The aim of this study is segmentation of the food Market based on this model. The study sample is 383 residents of Shiraz selected by cluster sampling and completed the
more » ... y's questionnaire. In order to identify clusters of customers, the reliability and validity of the standard questionnaire was evaluated based on the data collected and eventually modified questionnaire with 19 dimensions and 55 items was obtained. Then, based on the modified questionnaire clustering process in two steps was done and leads to the identification of three clusters of traditional, Unaware and rational in the city of Shiraz.
doi:10.22059/jibm.2016.58697 doaj:d676898a0d35428bb3baff79b83c46ec fatcat:z4j2vdqrnvexnaljsrcyuc7zgu

Mycophenolic Acid Pharmacokinetics Early After Kidney Transplant

Nazanin Honarbakhsh, Mohammad Reza Rouini, Mahboob Lesan-Pezeshki, Mohammad Reza Javadi, Iman Karimzadeh, Niayesh Mohebbi, Kheirollah Gholami
2013 Experimental and Clinical Transplantation  
Objectives: To determine the mycophenolic acid pharmacokinetic profile early after transplant in Iranian kidney graft recipients. Materials and Methods: A cross-sectional study was performed during 6 months in 31 patients who recently had kidney transplant and received fixed doses of mycophenolate mofetil (2 g/d). The plasma levels of mycophenolic acid were determined by high performance liquid chromatography. Results: The mean first mycophenolic acid peak level was 10 ± 5 mg/L. The mean
more » ... nolic acid area under the curve was 26 ± 19 mgh/L and apparent clearance was 57 ± 55 L/h. The mycophenolic acid area under the curve values of only 8 patients (26%) were within the therapeutic range (30-60 mgh/L). The first, second, and third mycophenolic acid peak levels correlated significantly with mycophenolic acid area under the curve (P < .05). Mycophenolic acid concentration at 10 hours had the highest correlation with mycophenolic acid area under the curve (r=0.962; P < .05). No statistically significant differences were evident in the mean mycophenolic acid area under the curve between men and women. Conclusions: There was a high degree of variation between different patients in mycophenolic acid pharmacokinetics early after kidney transplant.
doi:10.6002/ect.2012.0094 pmid:23176542 fatcat:s2emc35zkne7jpqi2efxed36my

Distinction between features extracted using deep belief networks [article]

Mohammad Pezeshki, Sajjad Gholami, Ahmad Nickabadi
2014 arXiv   pre-print
Data representation is an important pre-processing step in many machine learning algorithms. There are a number of methods used for this task such as Deep Belief Networks (DBNs) and Discrete Fourier Transforms (DFTs). Since some of the features extracted using automated feature extraction methods may not always be related to a specific machine learning task, in this paper we propose two methods in order to make a distinction between extracted features based on their relevancy to the task. We
more » ... lied these two methods to a Deep Belief Network trained for a face recognition task.
arXiv:1312.6157v2 fatcat:7aie3fpbn5fxbhssayj2ytpt3u

Diagnostic Value of Ultrasonography and Magnetic Resonance Imaging in Ulnar Neuropathy at the Elbow

Hormoz Ayromlou, Mohammad K. Tarzamni, Mohammad Hossein Daghighi, Mohammad Zakaria Pezeshki, Mohammad Yazdchi, Elyar Sadeghi-Hokmabadi, Ehsan Sharifipour, Kamyar Ghabili
2012 ISRN Neurology  
Aim. To evaluate the diagnostic value of ultrasonography and magnetic resonance imaging (MRI) in patients with ulnar neuropathy at the elbow (UNE). Methods. We prospectively performed electrodiagnostic, ultrasonographic, and MRI studies in UNE patients and healthy controls. Three cross-sectional area (CSA) measurements of the ulnar nerve at multiple levels along the arm and maximum CSA(-max) were recorded. Results. The ulnar nerve CSA measurements were different between the UNE severity grades
more » ... P<0.05). CSA-max had the greatest sensitivity (93%) and specificity (68%). Moreover, CSA-max ≥10 mm2 defined the severe UNE cases (sensitivity/specificity: 82%/72%). In MRI, ulnar nerve hyperintensity had the greatest sensitivity (90%) and specificity (80%). Conclusion. Ultrasonography using CSA-max is sensitive and specific in UNE diagnosis and discriminating the severe UNE cases. Furthermore, MRI particularly targeting at increased signal of the ulnar nerve can be a useful diagnostic test of UNE.
doi:10.5402/2012/491892 pmid:22888452 pmcid:PMC3409534 fatcat:y2kgyiygnrafrhzdx6rjko5v2u

Moyamoya disease mimicking encephalitis

Maryam Khalesi, Masoud Pezeshki Rad, Abdolkarim Hamedi, Mohammad Hassan Aelami
2014 Iranian Journal of Medical Sciences  
Moyamoya disease is a rare vaso-occlusive illness with an unknown etiology characterized by stenosis of the internal carotid arteries with spontaneous development of a collateral vascular network. A 15-month-old girl was referred to the emergency ward of Imam Reza Hospital due to decreased level of consciousness, focal seizures and fever during the previous 24 hours with an impression of encephalitis. Physical examination revealed left side hemiparesis; however brain CT-Scan did not show any
more » ... nificant lesions. Initial therapy with vancomycin, ceftriaxone and acyclovir was administered. CSF analysis did not show any abnormality and the blood as well as CSF cultures results were negative. Brain MRI showed hyperintensity at right frontal and parietal regions, suggesting vascular lesion. Magnetic resonance angiography (MRA) showed bilaterally multiple torsions in vessels at the basal ganglia consistent with moyamoya vessels. In all children exhibiting encephalitis, vascular events such as moyamoya disease should be considered. Brain MRI is a critical tool for this purpose. Common causes of encephalitis such as herpes simplex should also be ruled out.
pmid:25242851 pmcid:PMC4164900 fatcat:yd55l2vgi5cvdmgm524fsehzy4

Antibiofilm Activity of Methanol Extract of Rumex dentatus Against Pseudomonas aeruginosa

Maryam Pezeshki Najafabadi, Maryam Mohammadi-Sichani, Mohammad Javad Kazemi, Mohammad Sadegh Shirsalimian, Majid Tavakoli
2020 Biosis:Biological Systems  
In our previous study (Pezeshki Najafabadi et al., 2016) , in accordance with this one, PCR results and images were presented.  ... 
doi:10.37819/biosis.v1i1.44 fatcat:ds3tjs2gvfhztgbwwlowhtagvi

Towards End-to-End Speech Recognition with Deep Convolutional Neural Networks [article]

Ying Zhang, Mohammad Pezeshki, Philemon Brakel, Saizheng Zhang, Cesar Laurent Yoshua Bengio, Aaron Courville
2017 arXiv   pre-print
Convolutional Neural Networks (CNNs) are effective models for reducing spectral variations and modeling spectral correlations in acoustic features for automatic speech recognition (ASR). Hybrid speech recognition systems incorporating CNNs with Hidden Markov Models/Gaussian Mixture Models (HMMs/GMMs) have achieved the state-of-the-art in various benchmarks. Meanwhile, Connectionist Temporal Classification (CTC) with Recurrent Neural Networks (RNNs), which is proposed for labeling unsegmented
more » ... uences, makes it feasible to train an end-to-end speech recognition system instead of hybrid settings. However, RNNs are computationally expensive and sometimes difficult to train. In this paper, inspired by the advantages of both CNNs and the CTC approach, we propose an end-to-end speech framework for sequence labeling, by combining hierarchical CNNs with CTC directly without recurrent connections. By evaluating the approach on the TIMIT phoneme recognition task, we show that the proposed model is not only computationally efficient, but also competitive with the existing baseline systems. Moreover, we argue that CNNs have the capability to model temporal correlations with appropriate context information.
arXiv:1701.02720v1 fatcat:46c5bdvoofgmtfl33nlhysdzku

Vitamin D and its receptor polymorphisms: New possible prognostic biomarkers in leukemias

Seyed Mohammad Sadegh Pezeshki, Ali Amin Asnafi, Abbas Khosravi, Mohammad Shahjahani, Shirin Azizidoost, Saeid Shahrabi
2018 Oncology Reviews  
Several factors such as chromosomal translocations, gene mutations, and polymorphisms are involved in the pathogenesis of leukemia/lymphoma. Recently, the role of vitamin D (VD) and vitamin D receptor (VDR) polymorphisms in hematologic malignancies has been considered. In this review, we examine the possible role of VD levels, as well as VDR polymorphisms as prognostic biomarkers in leukemia/lymphoma. Relevant English language literature were searched and retrieved from Google Scholar search
more » ... ine (1985-2017). The following keywords were used: vitamin D, vitamin D receptor, leukemia, lymphoma, and polymorphism. Increased serum levels of VD in patients with leukemia are associated with a better prognosis. However, low VD levels are associated with a poor prognosis, and VDR polymorphisms in various leukemias can have prognostic value. VD biomarker can be regarded as a potential prognostic factor for a number of leukemias, including acute myeloblastic leukemia (AML), chronic lymphoblastic leukemia (CLL), and diffuse large B-cell lymphoma (DLBCL). There is a significant relationship between different polymorphisms of VDR (including Taq I and Fok I) with several leukemia types such as ALL and AML, which may have prognostic value.
doi:10.4081/oncol.2018.366 pmid:30405894 pmcid:PMC6199555 fatcat:aitcxsfljbaphprtdcla2qcz4m

Model of Cholera Forecasting Using Artificial Neural Network in Chabahar City, Iran

Zahra Pezeshki, Mohammad Tafazzoli-Shadpour, Isar Nejadgholi, Ali Mansourian, Mohammad Rahbar
2016 International Journal of Enteric Pathogens  
Footnote Authors' Contribution:Literature review, data collection, analysis and drafting: Zahra Pezeshki; editing and preparation: Mohammad Tafazzoli-Shadpour; technical advice for neural network, preparing  ...  Acknowledgments The authors would like to express gratitude to the personnel of Chabahar district health center, especially Mohammad Ali Abouzari and Faiz Mohammad Atash-Bahar for their valuable assistance  ... 
doi:10.17795/ijep31445 fatcat:5zb3c6bz4fdcdd2oemcmcqr2cq

What is the relationship between family physicians' job satisfaction associated and their educational performance?

Hossein Jabbari, Mohammad Zakarria Pezeshki, Mohammad Naghavi-Behzad, Mohammad Asghari, Reza Piri, Fariba Bakhshian
2015 Journal of Analytical Research in Clinical Medicine  
doi:10.15171/jarcm.2015.005 fatcat:golb4iquv5dc3dkk33f2x3k6aq
« Previous Showing results 1 — 15 out of 231 results