Sambaiga

I am Anthony Faustine, a Lead data scientist at the Centre of Intelligence Power (CIP) Eaton, Dublin. I have experience applying data science and machine learning techniques to business problems, with a track record in delivering and managing AI projects in academia and industries. Previously I was a Senior Industrial Analytics Researcher at Irish Manufacturing (IMR), Dublin, Ireland, where I developed data analytics solutions and tools to help industries and manufacturing design more effective solutions.

Between 2012 and 2017, I worked as an assistant lecturer at the University of Dodoma, Tanzania, where I was involved in several research projects within ICT for development (ICT4D). In 2017, I joined IDLab, imec research group of the University of Ghent in Belgium as a Machine learning researcher. From 1st February 2020 to 30th March 2021, I worked as a Data Scientist and Machine learning researcher at CeADAR (UCD), Dublin, Ireland, where I devised and implemented data analytics and AI technical solutions for business problems.

I am also PhD research Fellow at Further Energy and Environment Research Laboratory (FEELab), at ITI/LARSyS, Técnico Lisboa in Portugal under Prof Nuno Jardim Nunes and Dr. Lucas Pereira supervision. I am currently doing a Masters progam; MSc In Leadership, Innovation And Technology offered by TU Dublin (Grangegorman Campus) in partnership with Technology Ireland ICT.

My research focuses on developing robust machine-learning algorithms for real-world industrial applications, focusing on the following key areas.

Machine learning and signal processing for time series analysis focusing on Energy-disaggregation, Power demand forecasting, Anomaly detection and predictive maintenance.
Computer vision techniques for machine vision focusing on vision anomaly detection, shape matching and visual inspection.
Artificial Intelligence and Earth Observations(EO) for sustainability focusing on sustainable energy, agriculture and natural resources management.
Data management focuses on specification, strategies, best practices and guidelines for AI dataset development.

I also maintain a repository with resources on learning and applying AI for practical problems.

selected publications

FPSeq2Q

FPSeq2Q: Fully Parameterized Sequence to Quantile Regression for Net-Load Forecasting with Uncertainty Estimates

Faustine, Anthony, and Pereira, Lucas

IEEE Transactions on Smart Grid 2022

Abs PDF Code

The increased penetration of Renewable Energy Sources (RES) as part of a decentralized and distributed power system makes net-load forecasting a critical component in the planning and operation of power systems. However, compared to the transmission level, producing accurate short-term netload forecasts at the distribution level is complex due to the small number of consumers. Moreover, owing to the stochastic nature of RES, it is necessary to quantify the uncertainty of the forecasted net-load at any given time, which is critical for the real-world decision process. This work presents parameterized deep quantile regression for short-term probabilistic net-load forecasting at the distribution level. To be precise, we use a Deep Neural Network (DNN) to learn both the quantile fractions and quantile values of the quantile function. Furthermore, we propose a scoring metric that reflects the trade-off between predictive uncertainty performance and forecast accuracy. We evaluate the proposed techniques on historical real-world data from a lowvoltage distribution substation and further assess its robustness when applied in real-time. The experiment’s outcomes show that the resulting forecasts from our approach are well-calibrated and provide a desirable trade-off between forecasting accuracies and predictive uncertainty performance that are very robust even when applied in real-time.
AWRGNILM

Adaptive Weighted Recurrence Graphs for Appliance Recognition in Non-Intrusive Load Monitoring

Faustine, Anthony, Pereira, Lucas, and Klemenjak, Christoph

IEEE Transactions on Smart Grid 2021

Abs PDF Code

To this day, hyperparameter tuning remains a cumbersome task in Non-Intrusive Load Monitoring (NILM) research, as researchers and practitioners are forced to invest a considerable amount of time in this task. This paper proposes adaptive weighted recurrence graph blocks (AWRG) for appliance feature representation in event-based NILM. An AWRG block can be combined with traditional deep neural network architectures such as Convolutional Neural Networks for appliance recognition. Our approach transforms one cycle per activation current into an weighted recurrence graph and treats the associated hyper-parameters as learn-able parameters. We evaluate our technique on two energy datasets, the industrial dataset LILACD and the residential PLAID dataset. The outcome of our experiments shows that transforming current waveforms into weighted recurrence graphs provides a better feature representation and thus, improved classification results. It is concluded that our approach can guarantee uniqueness of appliance features, leading to enhanced generalisation abilities when compared to the widely researched V-I image features. Furthermore, we show that the initialisation parameters of the AWRG’s have a significant impact on the performance and training convergence.