Papers by mohammad norouzi
Applied Surface Science, 2016
Abstract In this study, the physico-chemical effects occasioned by the cold plasma discharge (CPD... more Abstract In this study, the physico-chemical effects occasioned by the cold plasma discharge (CPD) on the photo-decolorization of Reactive Orange 16 (RO16) by 3D fabrics (spacer fabrics) loaded with ZnO:TiO2 nano-photocatalysts (nphs) were optimized via response surface methodology (RSM). CPD was employed to improve the surface characteristics of the spacer fabrics for nphs loading. Surface morphology and color variation were studied utilizing scanning electron microscopy (SEM) and CIE-Lab system, respectively. The effect of CPD on the wetting ability of the spacer fabrics was examined using dynamic adsorption measurement (DAM). Also, X-ray fluorescence (XRF) was utilized to investigate the durability of the nphs on the spacer fabrics. All the experiments were implemented in a Box–Behnken design (BBD) with three independent variables (CPD treatment time, dye concentration and irradiation time) in order to optimize the decolorization of RO16. The anticipated values of the decolorization efficiency were found to be in excellent agreement with the experimental values (R2 = 0.9996, Adjusted R2 = 0.9992). The kinetic analysis demonstrated that the photocatalytic decolorization followed the Langmuir–Hinshelwood kinetic model. In conclusion, this heterogeneous photocatalytic process is capable of decolorizing and mineralizing azoic reactive dye in textile wastewater. Moreover, the results confirmed that RSM based on the BBD was a suitable method to optimize the operating conditions of RO16 degradation.
arXiv (Cornell University), Sep 1, 2016
A key problem in structured output prediction is direct optimization of the task reward function ... more A key problem in structured output prediction is direct optimization of the task reward function that matters for test evaluation. This paper presents a simple and computationally efficient approach to incorporate task reward into a maximum likelihood framework. By establishing a link between the log-likelihood and expected reward objectives, we show that an optimal regularized expected reward is achieved when the conditional distribution of the outputs given the inputs is proportional to their exponentiated scaled rewards. Accordingly, we present a framework to smooth the predictive probability of the outputs using their corresponding rewards. We optimize the conditional log-probability of augmented outputs that are sampled proportionally to their exponentiated scaled rewards. Experiments on neural sequence to sequence models for speech recognition and machine translation show notable improvements over a maximum likelihood baseline by using reward augmented maximum likelihood (RAML), where the rewards are defined as the negative edit distance between the outputs and the ground truth labels.
arXiv (Cornell University), Jan 5, 2015
We consider a quantum dot of MoS 2 which is made by decreasing the width of finite length of a Mo... more We consider a quantum dot of MoS 2 which is made by decreasing the width of finite length of a MoS 2 zigzag nanoribbon. The spin-dependent conductance and spin-polarization of the device are studied in presence (absent) of an external electric field perpendicular to the molybdenum plane by using the tight-binding non-equilibrium Green's function method. It is shown that in absence of the electric field, the deformation of ribbon structure causes spinsplitting in the dot and non-prefect spin filtering is seen. Therefore, the technique could be used for designing spin-dependent devices of MoS 2. Also, we show that the device could behave as a prefect spin filter and spin inverter under applying an external electric.
ArXiv, 2020
To select effective actions in complex environments, intelligent agents need to generalize from p... more To select effective actions in complex environments, intelligent agents need to generalize from past experience. World models can represent knowledge about the environment to facilitate such generalization. While learning world models from high-dimensional sensory inputs is becoming feasible through deep learning, there are many potential ways for deriving behaviors from them. We present Dreamer, a reinforcement learning agent that solves long-horizon tasks purely by latent imagination. We efficiently learn behaviors by backpropagating analytic gradients of learned state values through trajectories imagined in the compact state space of a learned world model. On 20 challenging visual control tasks, Dreamer exceeds existing approaches in data-efficiency, computation time, and final performance.
ArXiv, 2019
We consider the problem of learning from sparse and underspecified rewards, where an agent receiv... more We consider the problem of learning from sparse and underspecified rewards, where an agent receives a complex input, such as a natural language instruction, and needs to generate a complex response, such as an action sequence, while only receiving binary success-failure feedback. Such success-failure rewards are often underspecified: they do not distinguish between purposeful and accidental success. Generalization from underspecified rewards hinges on discounting spurious trajectories that attain accidental success, while learning from sparse feedback requires effective exploration. We address exploration by using a mode covering direction of KL divergence to collect a diverse set of successful trajectories, followed by a mode seeking KL divergence to train a robust policy. We propose Meta Reward Learning (MeRL) to construct an auxiliary reward function that provides more refined feedback for learning. The parameters of the auxiliary reward function are optimized with respect to the...
3 Abstract: 2-Methylpyridinium trifluoromethanesulfonate ((2-MPyH)OTf) catalyzed synthesis of 1,5... more 3 Abstract: 2-Methylpyridinium trifluoromethanesulfonate ((2-MPyH)OTf) catalyzed synthesis of 1,5- benzodiazepine and quinoxaline derivatives in the reaction of between o-phenylenediamine and ketone (synthesis of 1,5-benzodiazepine derivatives) or 1,2-diketone (synthesis of quinoxaline derivatives) in high yields and short time reaction. The catalyst can be separated from the products by a change in the solvent. The catalyst is reusable.
Proceedings of the ACM International Conference on Supercomputing, 2019
A major task of parallelization with OpenMP is to decide where in a program to insert which OpenM... more A major task of parallelization with OpenMP is to decide where in a program to insert which OpenMP construct such that speedup is maximized and correctness is preserved. Another challenge is the classification of variables that appear in a construct according to their data-sharing semantics. Manual classification is tedious and error prone. Moreover, the choice of the data-sharing attribute can significantly affect performance. Grounded on the notion of parallel design patterns, we propose a method that identifies code regions to parallelize and selects appropriate OpenMP constructs for them. Also, we classify variables in the chosen constructs by analyzing data dependences that have been dynamically extracted from the program. Using our approach, we created OpenMP versions of 49 sequential benchmarks and compared them with the code produced by three state-of-the-art parallelization tools: Our codes are faster in most cases with average speedups relative to any of the three ranging from 1.8 to 2.7. Additionally, we automatically reclassified variables of OpenMP programs parallelized manually or with the help of these tools, improving their execution time by up to 29%.
RSC Advances, 2015
1-Methyl imidazolium tricyanomethanide {[HMIM]C(CN)3} as a novel nano molten salt (MS) was synthe... more 1-Methyl imidazolium tricyanomethanide {[HMIM]C(CN)3} as a novel nano molten salt (MS) was synthesized and applied for the synthesis of 1-amidoalkyl-2-naphthols and compared with tin dioxide (SnO2) nanoparticles.
Neural Computing and Applications, 2014
It is a critical step to choose visual features in object tracking. Most existing tracking approa... more It is a critical step to choose visual features in object tracking. Most existing tracking approaches adopt handcrafted features, which greatly depend on people's prior knowledge and easily become invalid in other conditions where the scene structures are different. On the contrary, we learn informative and discriminative features from image data of tracking scenes itself. Local receptive filters and weight sharing make the convolutional restricted Boltzmann machines (CRBM) suit for natural images. The CRBM is applied to model the distribution of image patches sampled from the first frame which shares same properties with other frames. Each hidden variable corresponding to one local filter can be viewed as a feature detector. Local connections to hidden variables and maxpooling strategy make the extracted features invariant to shifts and distortions. A simple naive Bayes classifier is used to separate object from background in feature space. We demonstrate the effectiveness and robustness of our tracking method in several challenging video sequences. Experimental results show that features automatically learned by CRBM are effective for object tracking.
ChemInform, 2014
ABSTRACT The Hantzsch three-component condensation reaction of various aromatic aldehydes, 1,3-di... more ABSTRACT The Hantzsch three-component condensation reaction of various aromatic aldehydes, 1,3-dione and aniline derivatives in the presence of 2-methylpyridinium trifluoromethanesulphonate ([2-MPyH]OTf) as green and highly efficient catalysts in water affords 1,8-dioxodecahydroacridine derivatives in good to excellent yields. This reaction has been carried out in the presence of 1 mol% of [2-MPyH]OTf at room temperature. The described novel synthesis method proposes several advantages of mild condition, short reaction times, high yields, simplicity and easy workup compared to the traditional method of synthesis. Graphical Abstract1,8-Dioxodecahydroacridine derivatives were synthesized in good to excellent yield using [2-MPyH]OTf as an ionic liquid catalyst.
The Journal of Supercomputing, 2014
Efficient resource discovery plays a vital role in the effective management of resources and appl... more Efficient resource discovery plays a vital role in the effective management of resources and applications in heterogeneous computing environments. Therefore, the knowledge of applications' behavior and resources' usage pattern improves resource discovery decisions. This knowledge can be provided for the resource discovery mechanism by cooperating with the load balancing mechanism. In this paper, we formulate their cooperation by considering some parameters that represent applications' behavior and resources' usage patterns and extract the relation between them to introduce a formula using mathematical methods. Further, the resource discovery mechanism uses the formula to predict resources' load before assigning them new processes and thus it prevents resource overloading which happens frequently in computing environments.
2009 IEEE Conference on Computer Vision and Pattern Recognition, 2009
In this paper we present a method for learning classspecific features for recognition. Recently a... more In this paper we present a method for learning classspecific features for recognition. Recently a greedy layerwise procedure was proposed to initialize weights of deep belief networks, by viewing each layer as a separate Restricted Boltzmann Machine (RBM). We develop the Convolutional RBM (C-RBM), a variant of the RBM model in which weights are shared to respect the spatial structure of images. This framework learns a set of features that can generate the images of a specific object class. Our feature extraction model is a four layer hierarchy of alternating filtering and maximum subsampling. We learn feature parameters of the first and third layers viewing them as separate C-RBMs. The outputs of our feature extraction hierarchy are then fed as input to a discriminative classifier. It is experimentally demonstrated that the extracted features are effective for object detection, using them to obtain performance comparable to the state-of-the-art on handwritten digit recognition and pedestrian detection.
Journal of Chemical Sciences, 2013
The Hantzsch three-component condensation reaction of various aromatic aldehydes, 1,3-dione and a... more The Hantzsch three-component condensation reaction of various aromatic aldehydes, 1,3-dione and aniline derivatives in the presence of 2-methylpyridinium trifluoromethanesulphonate ([2-MPyH]OTf) as green and highly efficient catalysts in water affords 1,8-dioxodecahydroacridine derivatives in good to excellent yields. This reaction has been carried out in the presence of 1 mol% of [2-MPyH]OTf at room temperature. The described novel synthesis method proposes several advantages of mild condition, short reaction times, high yields, simplicity and easy workup compared to the traditional method of synthesis.
This paper proposes a new technique based on lexicographic optimization and ε-constraint method t... more This paper proposes a new technique based on lexicographic optimization and ε-constraint method to solve the combined economic emission scheduling problem of hydrothermal systems comprising several equality constraints as well as non-equality ones. The hydrothermal scheduling problem is modeled as a multi-objective problem with two objective functions as fuel cost minimization and also pollutant emission minimization. After deriving the Pareto set, the most preferred solution is determined using a fuzzy satisfying method. The thermal plants are considered with valve point effect and emission level function. Besides, this paper has taken into consideration the multi-reservoir cascaded configuration of hydro units, while the relationship considered between water discharge rate and power generated through these hydro units is nonlinear. Also, there is a water transport delay between hydro units. The presented model is implemented on a sample test system comprising four cascaded hydro units and three thermal units to verify the efficiency of the proposed method. Furthermore, the proposed method is implemented on IEEE 118 bus test case. The results obtained from the simulation show the effectiveness of the presented technique in the case of fuel cost and emission output compared to other approaches recently used.
2013 IEEE Conference on Computer Vision and Pattern Recognition, 2013
A fundamental limitation of quantization techniques like the k-means clustering algorithm is the ... more A fundamental limitation of quantization techniques like the k-means clustering algorithm is the storage and runtime cost associated with the large numbers of clusters required to keep quantization errors small and model fidelity high. We develop new models with a compositional parameterization of cluster centers, so representational capacity increases super-linearly in the number of parameters. This allows one to effectively quantize data using billions or trillions of centers. We formulate two such models, Orthogonal k-means and Cartesian k-means. They are closely related to one another, to k-means, to methods for binary hash function optimization like ITQ [5], and to Product Quantization for vector quantization [7]. The models are tested on largescale ANN retrieval tasks (1M GIST, 1B SIFT features), and on codebook learning for object recognition (CIFAR-10).
INTERNATIONAL …, 2007
Palm oil solid wastes consist of cellulose, hemicellulose and lignin. In this study, a single sta... more Palm oil solid wastes consist of cellulose, hemicellulose and lignin. In this study, a single stage of acid hydrolysis process of palm oil empty fruit bunch (EFB) for production of fermentable sugar was carried out under moderate temperature (45°C) and ambient pressure. The effect of four different process variables such as solid size, HCl concentration, solid percentage and temperature were investigated. In addition, the effect of pretreatment with 0.5-1M NaOH solution in acid hydrolysis was also investigated. Smaller particles provided more surfaces for acid-solid contact and longer reaction time was necessary if the large solid particles were used. High acid concentration improved the reaction rate and sugar yield. Therefore, the sugar yield was found to be dependent on acid concentration and the employed temperature as well. For a reaction time of 40 minutes, 5 % EFB solid with 15, 20, 25 and 30 percent of HCl, EFB lignocellulose fibers conversion of 36, 60, 65 and 80 % were achieved, respectively. The sugar concentration in acid hydrolysis of the pretreated fibers with 0.5 M sodium hydroxide solution resulted in 35 % more sugar.
Journal of Nanoparticle Research, 2010
This study attempted to synthesize onedimensional (1D) coaxial nanotubes of Fe 2 O 3 based on car... more This study attempted to synthesize onedimensional (1D) coaxial nanotubes of Fe 2 O 3 based on carbon nanotubes (CNT@Fe 2 O 3) via atomic layer deposition (ALD) using ferrocene and oxygen as precursors. Results disclosed that undoped CNTs were suitable for the ALD of Fe 2 O 3 (ALD-Fe 2 O 3) only if they were chemically functionalized, due to their inert surface nature. It was further demonstrated that the effects of both covalent and non-covalent methodologies were limited in functionalizing undoped CNTs, leading to random and non-uniform deposition of Fe 2 O 3. In sharp contrast, it was found that, as an alternative, nitrogen-doped CNTs (N-CNTs) contributed uniform and tunable ALD-Fe 2 O 3 , due to their active surface nature induced by incorporated N atoms. Consequently, various 1D heterostructural coaxial nanotubes were obtained with well-controlled growth of Fe 2 O 3 on N-CNTs. For a better understanding, the underlying mechanisms were explored based on different N-doping configurations. In addition, high-resolution transmission electron microscopy and X-ray diffraction jointly demonstrated that as-deposited Fe 2 O 3 is single-phase crystalline a-Fe 2 O 3 (hema-tite). The as-synthesized heterostructural coaxial nanotubes of CNT@Fe 2 O 3 may find great potential applications in photocatalysis, gas-sensing, and magnetic fields. Keywords Atomic layer deposition Á Iron oxide Á Carbon nanotubes Á Nitrogen-doping Á Coaxial nanostructures Á Nanocomposites
IEEE Transactions on Pattern Analysis and Machine Intelligence, 2014
There has been growing interest in mapping image data onto compact binary codes for fast near nei... more There has been growing interest in mapping image data onto compact binary codes for fast near neighbor search in vision applications. Although binary codes are motivated by their use as direct indices (addresses) into a hash table, codes longer than 32 bits are not being used in this way, as it was thought to be ineffective. We introduce a rigorous way to build multiple hash tables on binary code substrings that enables exact K-nearest neighbor search in Hamming space. The algorithm is straightforward to implement, storage efficient, and it has sub-linear run-time behavior for uniformly distributed codes. Empirical results show dramatic speed-ups over a linear scan baseline and for datasets with up to one billion items, 64-or 128-bit codes, and search radii up to almost 25 bits.
Comptes Rendus Chimie, 2014
Uploads
Papers by mohammad norouzi