A5010565726- Advanced Chemical Sensor Technologies
- Time Series Analysis and Forecasting
- Spectroscopy and Chemometric Analyses
- Anomaly Detection Techniques and Applications
- Music and Audio Processing
- Meat and Animal Product Quality
- Sentiment Analysis and Opinion Mining
- Spectroscopy Techniques in Biomedical and Chemical Research
- Stock Market Forecasting Methods
- Advanced Image and Video Retrieval Techniques
- Complex Systems and Time Series Analysis
- Advanced Clustering Algorithms Research
- Traffic Prediction and Management Techniques
- Face and Expression Recognition
University College Dublin
2020-2023
Institute for Development and Research in Banking Technology
2018-2020
University of Hyderabad
2018
The UEA Multivariate Time Series Classification (MTSC) archive released in 2018 provides an opportunity to evaluate many existing time series classifiers on the MTSC task. Nevertheless, although new TSC approaches were proposed recently, a comprehensive overview and empirical evaluation of techniques for task is currently missing from literature. In this work, we investigate state-of-the-art multivariate classification using benchmark. We compare recent methods originally developed...
The state-of-the-art in time series classification has come a long way, from the 1NN-DTW algorithm to ROCKET family of classifiers. However, current fast-paced development new classifiers, taking step back and performing simple baseline checks is essential. These are often overlooked, as researchers focused on establishing results, developing scalable algorithms, making models explainable. Nevertheless, there many datasets that look like at first glance, but classic algorithms such tabular...
In this study, we proposed two ensembled Convolutional Neural Network architectures viz. (CNNcuPSONN) and CNN-PNN, where cuPSONN is a CUDA enabled particle swarm optimization optimized neural network PNN the probabilistic network. We compared their performance with that of standalone PNN. The techniques are invoked after distributed Doc2Vec was employed on movie review dataset size 992 MB, to get paragraph embeddings dataset. Apache Spark framework used embeddings. Here, considered CNN as...
In April 2022, the Vistamilk SFI Research Centre organized second edition of "International Workshop on Spectroscopy and Chemometrics - Applications in Food Agriculture". Within this event, a data challenge was among participants workshop. Such competition aimed at developing prediction model to discriminate dairy cows' diet based milk spectral information collected mid-infrared region. fact, development an accurate reliable discriminant for can provide important authentication tools...
Accuracy is a key focus of current work in time series classification. However, speed and data reduction many applications equally important, especially when the scale storage requirements increase rapidly. Current MTSC algorithms need hundreds compute hours to complete training prediction. This due nature multivariate data, which grows with number series, their length channels. In applications, not all channels are useful for classification task; hence we require methods that can...