Munzali Alhassan

Munzali Alhassan ƒ - #- ï LinkedIn § Github Education Kano University of Science and Technology, Wudil 2016 - 2022 Bachelor of Electrical Engineering Technical Skills Languages: Python, R, SQL Frameworks: Pandas, Numpy, Tensorflow, Pytorch, Scikit-Learn, Seaborn Tools: PowerBI, Excel, PowerPoint Platform: HuggingFace, Pycharm, VSCode, Jupyter Notebook Projects UmojaHack Africa 2022: Faulty Air Quality Sensor Challenge Source Code • Developed a classification model to detect offset faults in air quality sensing devices deployed across Uganda. • Processed, cleaned and feature engineered a dataset of 300,000 training readings and 100,000 testing readings, including features like PM2.5 sensor readings, temperature, and humidity. • Utilized Gradient Boosting for model training and achieved an F1 score of 0.9169 on the validation set and 0.9155 on the training set. End-to-end Rice Disease Classification using Pretrained Mobilenet Source Code • Utilized the MobileNet pretrained model for feature extraction and classification of rice disease images. • Preprocessed a dataset of rice disease images, including data augmentation to improve model robustness. • Fine-tuned the MobileNet model to achieve high accuracy in classifying various rice diseases. • Deployed the trained model to Hugging Face Spaces, providing an interactive interface for users to upload images and receive classification results. Quora Insincere text questions Classification Source Code • Preprocessed and cleaned the text data, including tokenization, stopword removal, and padding sequences for uniform input size. • Employed embedding layers to convert text data into dense vectors suitable for neural network processing. • Designed and implemented a neural network with dense layers to classify questions as sincere or insincere. • Achieved a training accuracy of 0.89 and a validation accuracy of 0.86 and an AUC scores of 0.95 on the training set and 0.93 on the validation set. Work Experience Arewa Data Science Academy Feb 2024 – June 2024 Deep Learning Fellowship Remote • Conducted in-depth research and analysis on various deep learning models and algorithms. • Utilized Pytorch to developed and implemented machine learning and deep learning projects, including image and text classification tasks. • Collaborated with a team of data scientists and engineers to enhance the performance and accuracy of deep learning models. • Presented findings and project outcomes to peers and mentors, receiving constructive feedback and accolades. • Mentored junior fellows and provided guidance on project development, coding practices, and algorithm optimization. Freelance Data Scientist Upwork • Successfully processed and managed the pretrained Fooocus stable diffusion model. • Collaborated with the client to meet specific deployment requirements and resolve any issues promptly. • Ensured seamless integration and operation of the system and finally deployed to the HuggingFace space. • Received excellent feedback from the client for the quality and efficiency of the work. Mar 2024 Remote Certificates/Courses Deeplearning.ai Advanced Tensorflow Developer Specialization Coursera Deeplearning.ai Tensorflow Developer Specialization Coursera Deeplearning.ai Mathematics for Machine Learning Specialization Coursera Google Data analytics Specialization Coursera Prompt Engineering for ChatGPT Coursera Machine Learning Specialization Coursera IBM Machine Learning Specialization Coursera Learning How to Learn: Powerful mental tools to help you master tough subjects Coursera Professional Certification Google Certified - Tensorflow Developer (Certification ID:-) Issued by : Google Date Issued : Jun. 2023 Professional Membership Data Scientist Network - DSN ID : DSN/2020/13154 Membership : State Assistant Lead Nigerian Society of Engineers - NSE ID : G75321 Membership : Graduate Member