Close

João Pereira

JP

DOWNLOAD CV
About Me

I am currently a PDEng Trainee in Data Science at Eindhoven University of Technology (TU/e), where I develop data science projects for industry. Most of the time I do machine learning and I am broadly interested in the fields of data science and artificial intelligence.

My interests consist of applying cutting-edge AI techniques to solve difficult problems raised in the context of real-world applications. I have been working on AI applications involving time series, text, images, and videos, in the context of areas like energy, healthcare, manufacturing, and AI for social good.

Prior to joining TU/e as a PDEng trainee, I completed a M.Sc. degree in Electrical and Computer Engineering at Instituto Superior Técnico (University of Lisbon), where I did my master thesis on deep learning. Meanwhile, I developed applied research on deep learning within the Signal and Image Processing Group at the Institute for Systems and Robotics (ISR-Lisbon).

If you want to know more about my work I invite you to check out the Projects section. Please feel free to discuss and share any comments about them with me ;)

Experience
Logo TUE
PDEng Trainee in Data Science
Eindhoven University of Technology

I have been developing data science projects for the industry. Check out the project section.

Logo LxMLS
Monitor
9th Lisbon Machine Learning School

Part of the organizing team of LxMLS'19 as a monitor. I had responsibilities in the organization of the school and helped the students solving the exercises during the lab sessions, implementing machine learning algorithms such as naive Bayes, hidden Markov models, conditional random fields, recurrent neural networks, and reinforcement learning. [Website]

Logo ISR
Researcher
Institute for Systems and Robotics

Developed applied research within the Signal and Image Processing Group on deep learning for time series data, in the context of energy and healthcare applications.

In particular, my research was in the astonishing area of deep learning and its applications to sequential data. Specifically, it was focused on investigating how deep generative models (such as variational autoencoders), recurrent neural networks and attention mechanisms can be successfuly used for time series data modeling and anomaly detection.

Logo EDP
Intern
EDP Group

Conducted a machine learning project on load time series forecasting.

Logo EDP
Intern
EDP Group

Designed and projected medium/low voltage networks.

Projects
Thesis Icon

M.Sc. Thesis

"Unsupervised Anomaly Detection in Time Series Data Using Deep Learning"

READ MORE

Logo UNICRI

READ MORE

Logo UNICRI

Solar potential assessment from satellite images

Image processing (segmentation and classification)

READ MORE

Logo ASML

READ MORE

Logo TE

READ MORE

Logo DOW

READ MORE

Logo EDP

READ MORE

Image ECG

READ MORE

Image Brabant Water

READ MORE

Publications

Logo Thesis

M.Sc. Thesis, Deep learning for time series data

Proposes a general, unsupervised, and scalable, approach for anomaly detection in time series data using a cutting-edge deep learning approach.

J. Pereira, Unsupervised Anomaly Detection in Time Series Data Using Deep Learning, Instituto Superior Técnico [Thesis, Slides, BibTeX].


Anomaly Detection in Solar Energy Generation Data

Anomaly detection in solar generation time series using a variational autoencoder model and recurrent neural networks.

Introduced the Variational Self-Attention Mechanism.

J. Pereira, M. Silveira, Unsupervised Anomaly Detection in Energy Time Series Data using Variational Recurrent Autoencoders with Attention

17th IEEE International Conference on Machine Learning and Applications (ICMLA'18), Orlando, Florida, USA, Oral Presentation [Paper, Slides, BibTeX]


Deep Representation learning and anomaly detection

Learning representations of ECG time series for anomaly detection.

J. Pereira, M. Silveira, Learning Representations from Healthcare Time Series Data for Unsupervised Anomaly Detection

2019 IEEE International Conference on Big Data and Smart Computing (BigComp'19), Kyoto, Japan, Oral Presentation [Paper, Slides, BibTeX].


Deep Representation learning and anomaly detection

Proposes a variational recurrent autoencoder for learning representations of electrocardiogram (ECG) sequences. Downstream task is anomaly detection.

J. Pereira, M. Silveira, Unsupervised Representation Learning and Anomaly Detection in ECG Sequences

International Journal of Data Mining and Bioinformatics (IJDMB) [Paper, BibTeX]


Logo Google Scholar

Education
Logo IST
Integrated Master in Electrical and Computer Engineering
Instituto Superior Técnico

Topics: Artificial Intelligence, Machine Learning, Digital Signal Processing, Image Processing and Computer Vision, Robotics, Control Systems

Logo UCLouvain
Exchange Student within Erasmus+ Program
Université catholique de Louvain

Courses: Machine Learning, Image Processing and Vision, Instrumentation and Sensors, Electric Power Systems.

Get in Touch!

mail [at] joao-pereira [dot] pt