Artificial Intelligence Portfolio

Table of Contents

• About Me
• Machine Learning Projects
  • Genetic Syndrome Classification with Embeddings
  • Avocado Price Analysis and Forecasting
  • Exploratory Analysis of Airline Flights
  • Customer Churn Prediction
  • GDP vs. ACBR Linear Regression
• Deep Learning Projects
  • Diabetes Prediction in Pima Indians
  • Bike Detection with Detectron2
  • Marine Animal Detection with YOLOv5
  • GYMBUDDY - AI Fitness Assistant
  • DQN - CartPole with OpenAI Gym
  • Q-Learning - Mountain Car with OpenAI Gym
  • Marine Garbage Detection with YOLOv9
• Contact

---

About Me

Hello! I’m Henrique, a final-year Systems Analysis and Development student at FIAP. My passion for Machine Learning and Deep Learning has led me to develop numerous projects in these fields. My portfolio reflects my dedication to applying AI and data science to solve real-world challenges. From customer churn prediction to advanced object detection, my work demonstrates a commitment to leveraging technology for impactful solutions. I look forward to sharing my projects and exploring opportunities to drive meaningful advancements in AI.

---

Machine Learning Projects

🧬 Genetic Syndrome Classification with Embeddings

Developed a machine learning pipeline to classify genetic syndromes based on 320-dimensional image embeddings. The project involved data preprocessing, exploratory analysis, t-SNE visualization, K-Nearest Neighbors classification, and performance evaluation using multiple metrics.

• Technologies Used: Python, Scikit-learn, NumPy, Matplotlib, Seaborn
• Repository: GitHub - Genetic Syndrome Classification
• 

🥑 Avocado Price Analysis and Forecasting

Conducted exploratory data analysis and forecasting models for avocado prices using EDA techniques. Developed predictive models to assist in price forecasting, helping companies optimize stock and sales strategies.

• Technologies Used: Python, Pandas, Scikit-learn, Matplotlib
• Repository: GitHub - Avocado Prices EDA & Model Training
• 

✈️ Exploratory Analysis of Airline Flights

Extensive exploratory analysis of airline flight data to predict fare prices. Processed and analyzed large flight datasets to identify insights on fare fluctuations, aiding in market forecasting.

• Technologies Used: Python, Pandas, Seaborn, Matplotlib
• Repository: GitHub - EDA & Airline Fare Model Training
• 

🏠 Customer Churn Prediction

Implemented algorithms such as KNN, SVC, RandomForestClassifier, and Logistic Regression to predict customer churn. Built robust models to help companies retain clients by predicting churn risks.

• Technologies Used: Python, Scikit-learn, Pandas, Matplotlib
• Repository: GitHub - Classification Exercise
• 

📈 GDP vs. ACBR Linear Regression

Analyzed economic indicators using Linear Regression. Developed models to forecast economic trends, contributing to data-driven decision-making.

• Technologies Used: Python, Scikit-learn, Pandas, Matplotlib
• Repository: GitHub - GDP vs. ACBR Linear Regression
• 

---

Deep Learning Projects

🧬 Diabetes Prediction in Pima Indians

Built a neural network model using Keras to predict diabetes occurrence in the Pima Indian population. This project involved constructing and training a deep neural network to support early diagnosis.

• Technologies Used: Python, Keras, TensorFlow
• Repository: GitHub - Pima Indians Diabetes Prediction
• 

🚴 Porto Seguro Challenge - Bike Detection with Detectron2

Implemented an object detection system to identify bikes using Detectron2 and the COCO dataset. The model was deployed on AWS for cloud-based training and monitoring, aiming to streamline insurance processes with real-time detection.

• Technologies Used: Python, Detectron2, COCO
• Repository: GitHub - Bike Detection Project
• 

🦈 Marine Animal Detection with YOLOv5

Developed an object detection model using YOLOv5 to identify marine animals, such as sharks, fish, and rays. This project contributes to marine research by supporting species monitoring and conservation efforts.

• Technologies Used: Python, YOLOv5
• Repository: GitHub - Shark Detection with YOLOv5
• 

🤖 GYMBUDDY - AI Fitness Assistant

Created an innovative fitness assistant using Keras and the GPT-3 API. This project integrates AI-based conversational agents to provide personalized fitness guidance to users.

• Technologies Used: Python, Keras, GPT-3 API
• Repository: GitHub - GYMBUDDY AI
• 

🎮 DQN - CartPole with OpenAI Gym

Implemented a Deep Q-Learning model to solve the CartPole balancing problem in OpenAI Gym. This project explores reinforcement learning strategies for continuous control challenges.

• Technologies Used: Python, OpenAI Gym, TensorFlow
• Repository: GitHub - CartPole DQN
• 

🚗 Q-Learning - Mountain Car with OpenAI Gym

Developed a Q-Learning model to solve the Mountain Car problem in OpenAI Gym. This project addresses reinforcement learning techniques for solving complex optimization challenges.

• Technologies Used: Python, OpenAI Gym
• Repository: GitHub - Mountain Car Q-Learning
• 

🌊 Marine Garbage Detection with YOLOv9

Developed a marine garbage detection model using YOLOv9. The model was optimized for real-time detection on edge devices, contributing to environmental protection efforts.

• Technologies Used: Python, YOLOv9
• Repository: GitHub - Sea Garbage Detection with YOLOv9
• 

---

📞 Contact

• LinkedIn : Henrique Baptista
• GitHub : henriquebap
• Email : [email protected]

---