A5101892987- Topic Modeling
- Natural Language Processing Techniques
- Sentiment Analysis and Opinion Mining
- Speech Recognition and Synthesis
- Analytical Chemistry and Sensors
- Multimodal Machine Learning Applications
- Electrochemical Analysis and Applications
- Advanced Text Analysis Techniques
- Speech and Audio Processing
- Water Quality Monitoring and Analysis
- Privacy-Preserving Technologies in Data
- Adversarial Robustness in Machine Learning
- Gas Sensing Nanomaterials and Sensors
- Domain Adaptation and Few-Shot Learning
- Advanced Battery Technologies Research
- Explainable Artificial Intelligence (XAI)
- Vehicle License Plate Recognition
- EEG and Brain-Computer Interfaces
- Music and Audio Processing
- Sensor Technology and Measurement Systems
- Speech and dialogue systems
- Software Reliability and Analysis Research
- Currency Recognition and Detection
- Healthcare and Environmental Waste Management
- Advanced Graph Neural Networks
Singapore University of Technology and Design
2019-2024
Chitkara University
2023-2024
Nanyang Technological University
2023
Graphic Era University
2023
JECRC University
2021
National University of Singapore
2019-2020
L V Prasad Eye Institute
2020
Singapore Institute of Technology
2020
Instituto Politécnico Nacional
2020
University of Electronic Science and Technology of China
2019
This work presents modeling of temperature and temporal drift characteristics Al2O3-gate Ion-Sensitive Field-Effect Transistor (ISFET) performance enhancement ISFET-based pH sensor by using machine learning (ML) techniques. The behavioral macromodel ISFET is built Simulation Program with Integrated Circuit Emphasis (SPICE), which incorporates the dependent behavior electrochemical device parameters. SPICE for exported as a subcircuit block designing constant-voltage (CV) constant-current...
Summary The paper presents modeling and simulation of ion‐sensitive field‐effect transistor (ISFET)‐based pH sensor with temperature‐dependent behavioral macromodel proposes to compensate the temperature drift in using intelligent machine learning (ML) models. is built SPICE by introducing electrochemical parameters a metal‐oxide‐semiconductor (MOSFET) model simulate ISFET characteristics. We account for dependence semiconductor our increase its robustness. then exported as subcircuit...
Monitoring of pH is crucial for several chemical and biochemical processes. ISFET (Ion-Sensitive Field-Effect Transistors)-based sensors are promising candidates monitoring applications. However, devices prone to temporal temperature drifts, which severely affects the precision measurements. In this work, we collect experimental data drifts in an sensor formulate accurate SPICE macro model, incorporating both non-idealities. The developed model utilized generating training state-of-the-art...
Fine-tuning is widely used as the default algorithm for transfer learning from pre-trained models. Parameter inefficiency can however arise when, during learning, all parameters of a large model need to be updated individual downstream tasks. As number grows, fine-tuning prone overfitting and catastrophic forgetting. In addition, full become prohibitively expensive when many To mitigate this issue, parameter-efficient algorithms, such adapters prefix tuning, have been proposed way introduce...
Larger language models (LLMs) have taken the world by storm with their massive multi-tasking capabilities simply optimizing over a next-word prediction objective. With emergence of properties and encoded knowledge, risk LLMs producing harmful outputs increases, making them unfit for scalable deployment public. In this work, we propose new safety evaluation benchmark RED-EVAL that carries out red-teaming. We show even widely deployed are susceptible to Chain Utterances-based (CoU) prompting,...
Stock price prediction is one of the most complex machine learning problems. It depends on a large number factors which contribute to changes in supply and demand. In this paper, we propose stock analysis using based support vector machines (SVM), linear regression reinforcement learning. SVM are favored applications where text mining used for market prediction. SVMs can be both linearly non-linearly separable data sets. when separable, construct hyperplane feature space distinguish training...
In this paper, we investigate the importance of social network information compared to content in prediction a Twitter user’s occupational class. We show that tweets, profile descriptions follower/following community, and provide useful for classifying group. our study, extend an existing data set problem, achieve significantly better performance by using homophily has not been fully exploited previous work. analysis, found graph convolutional exploit homophily, can competitive on with just...
The construction of a sustainable smart society based on the concepts geosensitive equality through use Vedic Structure, AI, Blockchain, and IoT is proposed in this article. idea imagines where social justice, environmental sustainability, cultural preservation coexist harmony with economic prosperity. framework put up as blueprint for building harmonious all- encompassing society, focus civic participation teamwork. In paper, it suggested that emerging technologies like blockchain, IoT, AI...
This paper presents a new Machine Learning based temperature compensation technique for Ion-Sensitive Field-Effect Transistor (ISFET). The circuit models various electronic devices like MOSFET are available in commercial Technology Computer Aided Design (TCAD) tools such as LT-SPICE but no built-in model exists ISFET. Considering SiO <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</inf> the sensing film, an ISFET was created and simulations...
Semi-parametric Nearest Neighbor Language Models (kNN-LMs) have produced impressive gains over purely parametric LMs, by leveraging large-scale neighborhood retrieval external memory datastores. However, there has been little investigation into adapting such models for new domains. This work attempts to fill that gap and suggests the following approaches kNN-LMs — 1) underlying LM (using Adapters), 2) expanding an additional adaptation datastore, 3) weights (scores) of retrieved neighbors...
Neural speech synthesis, or text-to-speech (TTS), aims to transform a signal from the text domain domain. While developing TTS architectures that train and test on same set of speakers has seen significant improvements, out-of-domain speaker performance still faces enormous limitations. Domain adaptation new can be achieved by fine-tuning whole model for each domain, thus making it parameter-inefficient. This problem solved Adapters provide parameter-efficient alternative adaptation....
We propose Ruby Teaming, a method that improves on Rainbow Teaming by including memory cache as its third dimension. The dimension provides cues to the mutator yield better-quality prompts, both in terms of attack success rate (ASR) and quality diversity. prompt archive generated has an ASR 74%, which is 20% higher than baseline. In diversity, outperforms 6% 3% Shannon's Evenness Index (SEI) Simpson's Diversity (SDI), respectively.
Prompt Tuning has been largely successful as a parameter-efficient method of conditioning large-scale pre-trained language models to perform downstream tasks. Thus far, soft prompt tuning learns fixed set task-specific continuous vectors, i.e., tokens that remain static across the task samples. A prompt, however, may not generalize well diverse kinds inputs comprises. In order address this, we propose Vector-quantized Input-contextualized Prompts (VIP) an extension framework. VIP...