Wanderson souza
PhD, AI DEVELOPER & DATA SCIENTIST
Hi, I'm Wanderson.
I am a data scientist, Ph.D in Mechanical Engineering, Master and B.S. in Computer Science. I have advanced knowledge in the whole machine learning stack such as: Libraries (NumPy, Pandas, PySpark, scikit-learn, SciPy, NLTK, TensorFlow, Keras, PyTorch, Matplotlib), Container and virtual environments (Docker Container – VirtualEnv – Anaconda), Programming Language, development and versioning environment (Python, SQL, Git, Jupyter Notebook), DBMS (MySQL, MongoDB and PostgresSQL).
Currently I work as a data scientist at NeoPTO, a technology company focused on developing innovative solutions in the field of process automation for patent creation and maintenance using artificial intelligence. Such processes include the classification and automatic analysis of patents. We have recently developed a system in the information retrieval area that assists in the semantic search process for patents based on the evolution of trends passed by the user.
On a daily basis, I always try to update me and refine the maximum knowledge related to new architectures, performance improvements and network optimization during long hours of training. I am an open source software enthusiast and most productivity management tools. I intend to develop applications using machine learning and deep learning models, whether in recognition activities, natural language processing, or computer vision applied to embedded or non-embedded systems; I am also interested in researching new methods and strategies to improve network performance.
EXPERIENCE
Latest experiences
2022
Data Scientist
NEOPTO
I currently work as a data scientist at the NeoPTO Intellectual Property. We are developing a system in the information retrieval area that assists in the semantic search process for patents based on the evolution of trends passed by the user.
2021
Data Scientist
LACINA
I worked as a data scientist at the Applied Intelligence and Computation Lab (LACINA). We have developed a multilabel supervised system for hierarchical patent classification with more than 270,000 classes.
2020
Researcher and Machine Learning Engineer
GPICEEMA (INSTRUMENTATION AND CONTROL GROUP IN ENERGY AND ENVIRONMENT STUDY)
I worked as a machine learning engineer and researcher at the Solar Energy laboratory at the Center for Alternative and Renewable Energies at UFPB and I am part of the Instrumentation and Control Group in Energy and Environment Studies (GPICEEMA) developing solutions with computer vision.
2015-2019
Machine Learning Specialist
METROREC / CBTU-JP
I worked between 2015 and 2019 as a researcher and machine learning specialist in the partnership between UFPB and METROREC / CBTU-JP. Throughout this period, I developed the VisionBerry project, in which we developed a computer vision system to estimate the multidirectional flow of passengers in subway transportation using embedded devices.
Line Crossing
Raspberry Pi + Picamera
Validation
Line Crossing
2017-2019
Research and Professor
IFPB (FEDERAL INSTITUTE OF PARAÍBA)
I was a professor of computer science and scientific methodology, teaching classes on programming basics, advanced Excel, and operating systems for technical education classes.
EDUCATION
2015-2019
Doctor of Sciences
UFPB (FEDERAL UNIVERSITY OF PARAÍBA)
I am a D.Sc in Mechanical Engineering. Throughout this period I developed a passenger estimation system using deep neural networks. In October 2017, the results of the VisionBerry project earned me participation in the Empreenda Santander contest in the technological innovation category and I reached the semi-final.
2011-2013
Master of Science
2004-2009
Bachelor of Science
UEPB (STATE UNIVERSITY OF PARAÍBA)
B.S. in Computer Science at the State University of Paraíba
SKILLS
Development
- Experience with Python;
- Expertise on the full stack machine learning development libraries (NumPy, pandas, PySpark, matplotlib, scikit-image, SciPy, tensorflow, Keras and PyTorch) since 2015.
Data analysis
Data analysis is an essential component for those working with machine learning. I use linear, non-linear and logistic regression models to predict results using real data, such as estimating a country's GDP or even understanding the progress of a pandemic. I have several data science certificates MOOC's on Coursera.
Software engineering
I take use of design patterns, code review techniques, documentation, respecting the pep8 style guide and other important concepts to improve code readability and maintenance.
Proactivity
I'm a proactive person, compromised and responsible with my work. Attributes that I believe that all developers must have in order to be a quality professional.
Team work
I adapt well in teamwork. I use GitHub and GitLab for version control. I like to share the use of applications for productivity management.
High academic background
I have a high qualification, papers published in the area and teaching experience.
SPECIFIC SKILLS
-
MACHINE LEARNING MODELS
-
Supervised Learning
-
Linear and Non linear Regression
-
Logistic Regression
-
Support Vector Machine
-
Naive Bayes
-
Neural Networks
-
XGBoost
-
CatBoost
-
-
Unsupervised Learning
-
HDBSCAN
-
K-Means
-
LDA
-
-
-
DEEP LEARNING ARCHITECTURE
-
VGGNet
-
ShallowNet
-
GoogleNet
-
AlexNet
-
SqueezeNet
-
ResNet
-
RetinaNet
-
Faster R-CNN
-
SSD
-
Mask R-CNN
-
YOLO - Real-Time Object Detection
-
-
LIBRARIES AND PACKAGES
-
OpenCV
-
matplotlib
-
scikit-image
-
scikit-learn
-
NLTK
-
SciPy
-
Numpy
-
Pandas
-
PySpark
-
BeautifulSoap
-
Tensorflow
-
Mxnet
-
Keras
-
PyTorch
-
-
METHODS TO IMPROVE PREDICTIONS
-
Data Augmentation
-
Ensemble
-
Fine-tuning Networks with Transfer Learning
-
Advanced Optimization Methods (Adagrad, Adadelta, RMSprop, Adam, Nadam).
-
COURSES & CERTIFICATIONS
-
Introduction to Artificial Intelligence by IBM
-
Python for Data Science and AI by IBM
-
Machine Learning with Python by IBM
-
Introduction to Deep Learning & Neural Networks with Keras by IBM
-
Deep Neural Networks with PyTorch by IBM
-
Building Deep Learning Models with TensorFlow by IBM
-
AI Capstone Project with Deep Learning by IBM
-
Math for Machine Learning by AWS
-
Natural Language Processing by Kaggle
-
Natural Language Processing with Classification and Vector Spaces by deeplearning.ai
-
Improving Deep Neural Networks: Hyperparameter tuning, Regularization and Optimization by deeplearning.ai
-
Structuring Machine Learning Projects by deeplearning.ai
-
Exploring and Preparing your Data with BigQuery by Google Cloud
-
Scrum Fundamentals Certified by SCRUMstudy
