Haris Jafar

Hardworking and passionate Msc(AI) post-graduate with strong organizational skills eager to secure entry-level Machine Learning position. Ready to help the team achieve all its goals.

• Knife Classification Model Hyperparamter Tuning(Fundamentals of ML) - Experimented with different models in the tf library and hyperparameters to improve Mean Accuracy Precision from 0.61 to 0.81.
• Improving Classification through Self-supervised Learning(Applied Machine Learning)- Utilizing the SIIM-ISIC Melanoma Classification dataset, used baseline model of ResNet50, we used SimCLR for pretraining, and OneCycleLR as learning rate scheduler achieving Kaggle Public and Private scores of 80.93% and 82.73%.
• Vehicle Re-identification(ComputerVision&DeepLearning)- Applied different CNN models, data augmentations along-with hyperparameter tuning to improve the model's Mean Accuracy Precision.
• Murmur Detection using Machine Learning(AI&Health)- preprocessed the murmur dataset on Kaggle, applied Random Forest, XGBoost and other ML techniques and created a confusion matrix for model accuracy comparison.

• AWS Certified Cloud Practitioner
• Intro to AI- Coursera
• Python for Everybody- Udemy