Obokparo Esibe
Data Analyst | Power BI · SQL · Python | AI Data Quality
New to Contra
Obokparo is building their profile!
An end-to-end deep learning project exploring two approaches to AAPL stock prediction using 10 years of market data (2015–2024).
0
4
Built two image classification models on 20,000+ leaf images across 15 disease classes. First model: custom CNN achieving 95.5% accuracy. Second model: MobileNetV2 fine-tuned via transfer learning achieving 98.3% accuracy. Covered full deep learning pipeline — EDA, preprocessing, data augmentation, training, evaluation, and model comparison. Built with Python and PyTorch. Full code documented on GitHub.
0
9
Trained a Deep Q-Network (DQN) agent from scratch using Python and PyTorch to solve the CartPole balancing problem. Implemented epsilon-greedy exploration, experience replay buffer, and reward shaping. Achieved 6 out of 10 perfect 500-step evaluation scores with a mean score of 321.1. Built as part of the Microsoft ML Learning Pathway. Full code and evaluation results documented on GitHub.
0
9