-
Discrete Natural Evolution Strategies
Authors:
Ahmad Ayaz Amin
Abstract:
Natural evolution strategies are a class of approximate-gradient black-box optimizers that have been successfully used for continuous parameter spaces. In this paper, we derive NES algorithms for discrete parameter spaces and demonstrate their effectiveness in tasks involving discrete parameters.
Natural evolution strategies are a class of approximate-gradient black-box optimizers that have been successfully used for continuous parameter spaces. In this paper, we derive NES algorithms for discrete parameter spaces and demonstrate their effectiveness in tasks involving discrete parameters.
△ Less
Submitted 29 March, 2024;
originally announced April 2024.
-
An IoT Based Water-Logging Detection System: A Case Study of Dhaka
Authors:
Md Manirul Islam,
Md. Sadad Mahamud,
Umme Salsabil,
A. A. M. Mazharul Amin,
Samiul Haque Suman
Abstract:
With a large number of populations, many problems are rising rapidly in Dhaka, the capital city of Bangladesh. Water-logging is one of the major issues among them. Heavy rainfall, lack of awareness and poor maintenance causes bad sewerage system in the city. As a result, water is overflowed on the roads and sometimes it gets mixed with the drinking water. To overcome this problem, this paper reali…
▽ More
With a large number of populations, many problems are rising rapidly in Dhaka, the capital city of Bangladesh. Water-logging is one of the major issues among them. Heavy rainfall, lack of awareness and poor maintenance causes bad sewerage system in the city. As a result, water is overflowed on the roads and sometimes it gets mixed with the drinking water. To overcome this problem, this paper realizes the potential of using Internet of Things to combat water-logging in drainage pipes which are used to move wastes as well as rainwater away from the city. The proposed system will continuously monitor real time water level, water flow and gas level inside the drainage pipe. Moreover, all the monitoring data will be stored in the central database for graphical representation and further analysis. In addition to that if any emergency arises in the drainage system, an alert will be sent directly to the nearest maintenance office.
△ Less
Submitted 25 February, 2024;
originally announced March 2024.
-
Guided Sketch-Based Program Induction by Search Gradients
Authors:
Ahmad Ayaz Amin
Abstract:
Many tasks can be easily solved using machine learning techniques. However, some tasks cannot readily be solved using statistical models, requiring a symbolic approach instead. Program induction is one of the ways that such tasks can be solved by means of capturing an interpretable and generalizable algorithm through training. However, contemporary approaches to program induction are not sophistic…
▽ More
Many tasks can be easily solved using machine learning techniques. However, some tasks cannot readily be solved using statistical models, requiring a symbolic approach instead. Program induction is one of the ways that such tasks can be solved by means of capturing an interpretable and generalizable algorithm through training. However, contemporary approaches to program induction are not sophisticated enough to readily be applied to various types of tasks as they tend to be formulated as a single, all-encompassing model, usually parameterized by neural networks. In an attempt to make program induction a viable solution for many scenarios, we propose a framework for learning parameterized programs via search gradients using evolution strategies. This formulation departs from traditional program induction as it allows for the programmer to impart task-specific code to the program 'sketch', while also enjoying the benefits of accelerated learning through end-to-end gradient-based optimization.
△ Less
Submitted 10 February, 2024;
originally announced February 2024.
-
Statistical Parameterized Physics-Based Machine Learning Digital Twin Models for Laser Powder Bed Fusion Process
Authors:
Yangfan Li,
Satyajit Mojumder,
Ye Lu,
Abdullah Al Amin,
Jiachen Guo,
Xiaoyu Xie,
Wei Chen,
Gregory J. Wagner,
Jian Cao,
Wing Kam Liu
Abstract:
A digital twin (DT) is a virtual representation of physical process, products and/or systems that requires a high-fidelity computational model for continuous update through the integration of sensor data and user input. In the context of laser powder bed fusion (LPBF) additive manufacturing, a digital twin of the manufacturing process can offer predictions for the produced parts, diagnostics for m…
▽ More
A digital twin (DT) is a virtual representation of physical process, products and/or systems that requires a high-fidelity computational model for continuous update through the integration of sensor data and user input. In the context of laser powder bed fusion (LPBF) additive manufacturing, a digital twin of the manufacturing process can offer predictions for the produced parts, diagnostics for manufacturing defects, as well as control capabilities. This paper introduces a parameterized physics-based digital twin (PPB-DT) for the statistical predictions of LPBF metal additive manufacturing process. We accomplish this by creating a high-fidelity computational model that accurately represents the melt pool phenomena and subsequently calibrating and validating it through controlled experiments. In PPB-DT, a mechanistic reduced-order method-driven stochastic calibration process is introduced, which enables the statistical predictions of the melt pool geometries and the identification of defects such as lack-of-fusion porosity and surface roughness, specifically for diagnostic applications. Leveraging data derived from this physics-based model and experiments, we have trained a machine learning-based digital twin (PPB-ML-DT) model for predicting, monitoring, and controlling melt pool geometries. These proposed digital twin models can be employed for predictions, control, optimization, and quality assurance within the LPBF process, ultimately expediting product development and certification in LPBF-based metal additive manufacturing.
△ Less
Submitted 13 November, 2023;
originally announced November 2023.
-
Natural Evolution Strategies as a Black Box Estimator for Stochastic Variational Inference
Authors:
Ahmad Ayaz Amin
Abstract:
Stochastic variational inference and its derivatives in the form of variational autoencoders enjoy the ability to perform Bayesian inference on large datasets in an efficient manner. However, performing inference with a VAE requires a certain design choice (i.e. reparameterization trick) to allow unbiased and low variance gradient estimation, restricting the types of models that can be created. To…
▽ More
Stochastic variational inference and its derivatives in the form of variational autoencoders enjoy the ability to perform Bayesian inference on large datasets in an efficient manner. However, performing inference with a VAE requires a certain design choice (i.e. reparameterization trick) to allow unbiased and low variance gradient estimation, restricting the types of models that can be created. To overcome this challenge, an alternative estimator based on natural evolution strategies is proposed. This estimator does not make assumptions about the kind of distributions used, allowing for the creation of models that would otherwise not have been possible under the VAE framework.
△ Less
Submitted 15 August, 2023;
originally announced August 2023.
-
Stochastic additive manufacturing simulation: from experiment to surface roughness and porosity prediction
Authors:
Yangfan Li,
Ye Lu,
Abdullah Al Amin,
Wing Kam Liu
Abstract:
Deterministic computational modeling of laser powder bed fusion (LPBF) process fails to capture irregularities and roughness of the scan track, unless expensive powder-scale analysis is used. In this work we developed a stochastic computational modeling framework based on Markov Chain Monte Carlo (MCMC) capable of capturing the irregularities of LPBF scan. The model is calibrated against AFRL sing…
▽ More
Deterministic computational modeling of laser powder bed fusion (LPBF) process fails to capture irregularities and roughness of the scan track, unless expensive powder-scale analysis is used. In this work we developed a stochastic computational modeling framework based on Markov Chain Monte Carlo (MCMC) capable of capturing the irregularities of LPBF scan. The model is calibrated against AFRL single track scan data using a specially designed tensor decomposition method, i.e., Higher-Order Proper Generalized Decomposition (HOPGD) that relies on non-intrusive data learning and construction of reduced order surrogate models. Once calibrated, the stochastic model can be used to predict the roughness and porosity at part scale at a significantly reduced computational cost compared to detailed powder-scale deterministic simulations. The stochastic simulation predictions are validated against AFRL multi-layer and multitrack experiments and reported as more accurate when compared with regular deterministic simulation results.
△ Less
Submitted 2 August, 2022;
originally announced August 2022.
-
Using BERT Embeddings to Model Word Importance in Conversational Transcripts for Deaf and Hard of Hearing Users
Authors:
Akhter Al Amin,
Saad Hassan,
Cecilia O. Alm,
Matt Huenerfauth
Abstract:
Deaf and hard of hearing individuals regularly rely on captioning while watching live TV. Live TV captioning is evaluated by regulatory agencies using various caption evaluation metrics. However, caption evaluation metrics are often not informed by preferences of DHH users or how meaningful the captions are. There is a need to construct caption evaluation metrics that take the relative importance…
▽ More
Deaf and hard of hearing individuals regularly rely on captioning while watching live TV. Live TV captioning is evaluated by regulatory agencies using various caption evaluation metrics. However, caption evaluation metrics are often not informed by preferences of DHH users or how meaningful the captions are. There is a need to construct caption evaluation metrics that take the relative importance of words in a transcript into account. We conducted correlation analysis between two types of word embeddings and human-annotated labeled word-importance scores in existing corpus. We found that normalized contextualized word embeddings generated using BERT correlated better with manually annotated importance scores than word2vec-based word embeddings. We make available a pairing of word embeddings and their human-annotated importance scores. We also provide proof-of-concept utility by training word importance models, achieving an F1-score of 0.57 in the 6-class word importance classification task.
△ Less
Submitted 24 June, 2022;
originally announced June 2022.
-
A Disability Lens towards Biases in GPT-3 Generated Open-Ended Languages
Authors:
Akhter Al Amin,
Kazi Sinthia Kabir
Abstract:
Language models (LM) are becoming prevalent in many language-based application spaces globally. Although these LMs are improving our day-to-day interactions with digital products, concerns remain whether open-ended languages or text generated from these models reveal any biases toward a specific group of people, thereby risking the usability of a certain product. There is a need to identify whethe…
▽ More
Language models (LM) are becoming prevalent in many language-based application spaces globally. Although these LMs are improving our day-to-day interactions with digital products, concerns remain whether open-ended languages or text generated from these models reveal any biases toward a specific group of people, thereby risking the usability of a certain product. There is a need to identify whether these models possess bias to improve the fairness in these models. This gap motivates our ongoing work, where we measured the two aspects of bias in GPT-3 generated text through a disability lens.
△ Less
Submitted 23 June, 2022;
originally announced June 2022.
-
Performance Evaluation of Cooperative NOMA-based Improved Hybrid SWIPT Protocol
Authors:
Ahmed Al Amin,
Soo Young Shin
Abstract:
This study proposes the integration of a cooperative non-orthogonal multiple access (CNOMA) and improved hybrid simultaneous wireless information and power transfer (IHS SWIPT) protocol (termed as CNOMA-IHS) to enhance the spectral efficiency (SE) of a downlink (DL) CNOMA communication system. CNOMA-IHS scheme can enhance the ergodic sum capacity (ESC) and energy efficiency (EE) of DL CNOMA by tra…
▽ More
This study proposes the integration of a cooperative non-orthogonal multiple access (CNOMA) and improved hybrid simultaneous wireless information and power transfer (IHS SWIPT) protocol (termed as CNOMA-IHS) to enhance the spectral efficiency (SE) of a downlink (DL) CNOMA communication system. CNOMA-IHS scheme can enhance the ergodic sum capacity (ESC) and energy efficiency (EE) of DL CNOMA by transferring additional symbols towards the users and energize the relay operation as well without any additional resources (e.g., time slot/frequency/code). The analytical and simulation results indicate that the proposed CNOMA-IHS scheme outperforms other existing SWIPT-based schemes (e.g., CNOMA with hybrid SWIPT, CNOMA with power-splitting SWIPT, wireless-powered CNOMA, CNOMA with time switching SWIPT, and orthogonal multiple access with IHS SWIPT) in terms of the ESC. Moreover, the CNOMA-IHS scheme also enhances EE compared with other conventional TS-SWIPT-based schemes, which is also illustrated by the simulation results. In addition, the proposed CNOMA-IHS scheme with the considered EE optimization technique outplayed the proposed CNOMA-IHS scheme without EE optimization and other existing TS-SWIPT-based schemes in terms of EE.
△ Less
Submitted 20 June, 2021;
originally announced June 2021.
-
EPiT : A Software Testing Tool for Generation of Test Cases Automatically
Authors:
Rosziati Ibrahim,
Ammar Aminuddin Bani Amin,
Sapiee Jamel,
Jahari Abdul Wahab
Abstract:
Software test cases can be defined as a set of condition where a tester needs to test and determine that the System Under Test (SUT) satisfied with the expected result correctly. This paper discusses the optimization technique in generating cases automatically by using EpiT (Eclipse Plug-in Tool). EpiT is developed to optimize the generation of test cases from source code in order to reduce time u…
▽ More
Software test cases can be defined as a set of condition where a tester needs to test and determine that the System Under Test (SUT) satisfied with the expected result correctly. This paper discusses the optimization technique in generating cases automatically by using EpiT (Eclipse Plug-in Tool). EpiT is developed to optimize the generation of test cases from source code in order to reduce time used for conventional manually creating test cases. By using code smell functionality, EpiT helps to generate test cases automatically from Java programs by checking its line of code (LOC). The implementation of EpiT will also be presented based on several case studies conducted to show the optimization of the test cases generated. Based on the results presented, EpiT is proven to solve the problem for software tester to generate test case manually and check the optimization from the source code using code smell technique.
△ Less
Submitted 22 July, 2020;
originally announced July 2020.
-
Capacity improvement of non-orthogonal multiple access downlink transmission by orbital angular momentum based mode division multiple access
Authors:
Ahmed Al Amin,
Soo Young Shin
Abstract:
In this paper, non-orthogonal multiple access (NOMA) downlink transmission is integrated with orbital angular momentum (OAM) based mode division multiple access (MDMA), called NOMA-OAM-MDMA. Different OAM modes can generate different OAM waves for different superimposed signals. So, every OAM wave will be transmitted a superimposed signal from the base station to the cell center user (CCU) and cel…
▽ More
In this paper, non-orthogonal multiple access (NOMA) downlink transmission is integrated with orbital angular momentum (OAM) based mode division multiple access (MDMA), called NOMA-OAM-MDMA. Different OAM modes can generate different OAM waves for different superimposed signals. So, every OAM wave will be transmitted a superimposed signal from the base station to the cell center user (CCU) and cell edge user (CEU). In this way, multiple OAM waves with multiple superimposed signals will transmit from BS to CCU and CEU simultaneously to enhance the capacity of NOMA downlink transmission. Finally, the effectiveness of the proposed schemes over the existing scheme and conventional orthogonal multiple access based scheme are demonstrated through the result analysis.
△ Less
Submitted 1 November, 2019;
originally announced November 2019.
-
Channel Capacity Enhancement of SWIPT based CNOMA Downlink Transmission with Orbital Angular Momentum over Rician Fading Channel
Authors:
Ahmed Al Amin,
Soo Young Shin,
Bhaskara Narottama
Abstract:
In this paper, the integration of cooperative non-orthogonal multiple access (CNOMA) and power splitting based simultaneous wireless information and power transfer (SWIPT) is proposed to improve the spectral and energy efficiency. Moreover, different orbital angular momentum (OAM) based signals are also considered in this paper to transmit additional symbols to cell center user (CCU) and cell edge…
▽ More
In this paper, the integration of cooperative non-orthogonal multiple access (CNOMA) and power splitting based simultaneous wireless information and power transfer (SWIPT) is proposed to improve the spectral and energy efficiency. Moreover, different orbital angular momentum (OAM) based signals are also considered in this paper to transmit additional symbols to cell center user (CCU) and cell edge user (CEU). In the proposed technique, CCU is used as energy-constrained relay to forward signal to CEU. CCU harvests energy from the transmitted signal by the BS using the power splitting technique. The additional symbols are transmitted from the BS to CCU and CEU by utilizing different modes of OAM. This paper investigates the ergodic sum capacity (ESC) of the proposed technique along with analytical derivations over Rician fading channels. Finally, the performance in terms of ESC and energy efficiency over conventional technique compared to conventional multiple access is demonstrated by analysis and simulation.
△ Less
Submitted 20 October, 2019; v1 submitted 17 July, 2019;
originally announced July 2019.
-
Performance Enhancement of Hybrid SWIPT Protocol for Cooperative NOMA Downlink Transmission
Authors:
A. A. Amin,
M. B. Uddin,
S. Y. Shin
Abstract:
Time splitting and power splitting incorporating, a hybrid Simultaneous Wireless Information and Power Transfer (SWIPT) based cooperative Non-Orthogonal Multiple Access (CNOMA) protocol is considered in this paper. Cell center user of the CNOMA system acts as a relay to enhance the reliability of the cell edge user (CEU). SWIPT is considered to empower the relay operation to avoid the battery drai…
▽ More
Time splitting and power splitting incorporating, a hybrid Simultaneous Wireless Information and Power Transfer (SWIPT) based cooperative Non-Orthogonal Multiple Access (CNOMA) protocol is considered in this paper. Cell center user of the CNOMA system acts as a relay to enhance the reliability of the cell edge user (CEU). SWIPT is considered to empower the relay operation to avoid the battery draining issue. To enhance the system performance in terms of ergodic sum capacity (ESC) and outage probabilities (OP), an integration of CNOMA strategy and hybrid SWIPT protocol for the downlink (DL) transmission is proposed here. By utilizing the idle link of hybrid SWIPT protocol an enhanced hybrid SWIPT protocol is proposed here to enhance the performance of CNOMA DL transmission. Moreover, Maximal ratio combining is utilized as a diversity combining technique at CEU to enhance the performance as well. The performance of the proposed protocol is examined in terms of ergodic sum capacity, outage probabilities and energy efficiency. Finally, the analytical results are justified by the Monte-Carlo simulation. Numerical results demonstrate that the proposed protocol with effective CNOMA strategy achieves superior performance than HS-CNOMA with selection combining.
△ Less
Submitted 21 June, 2019;
originally announced June 2019.
-
Capacity Enhancement of Cooperative NOMA over Rician Fading Channels with Orbital Angular Momentum
Authors:
Ahmed Al Amin,
Bhaskara Narottama,
Soo Young Shin
Abstract:
This letter proposes the usage of orbital angular momentum (OAM) for cooperative non-orthogonal multiple access (CNOMA) to enhance sum capacity (SC) for the future cellular communication system. The proposed CNOMA-OAM scheme is analyzed and compared with other schemes, i.e., conventional CNOMA, conventional orthogonal multiple access (OMA) with OAM. The impact of the power allocation factor for OA…
▽ More
This letter proposes the usage of orbital angular momentum (OAM) for cooperative non-orthogonal multiple access (CNOMA) to enhance sum capacity (SC) for the future cellular communication system. The proposed CNOMA-OAM scheme is analyzed and compared with other schemes, i.e., conventional CNOMA, conventional orthogonal multiple access (OMA) with OAM. The impact of the power allocation factor for OAM beam over SC is also analyzed. The analytical result is justified by simulation results which demonstrate that the proposed CNOMA-OAM provides higher SC compared to other schemes.
△ Less
Submitted 21 June, 2019;
originally announced June 2019.
