CST 499 - Week 6

 Hey everyone,

    This week, I made solid progress in my fast.ai studies while also applying what I learned toward my project work. I covered several advanced topics including Stable Diffusion, matrix multiplication, mean shift clustering, backpropagation, MLPs, autoencoders, and the fast.ai learner framework. These topics helped me better understand both the theory and implementation of deep learning models.

    This week, I focused on strengthening the foundational components of my project. I worked on understanding and implementing core neural network concepts, particularly backpropagation and matrix operations, which are essential for building and debugging models. I also explored the fast.ai learner framework and began setting up a structured training pipeline for my project. Additionally, I experimented with autoencoders to better understand representation learning, which may be useful for my project depending on how I structure my model.

    Next week, I plan to start integrating these concepts into a more complete project workflow. My goal is to build and train a working model using the fast.ai framework, including proper data preprocessing, model training, and evaluation. I also want to experiment further with model performance by applying normalization and initialization techniques. If time allows, I plan to revisit difficult topics like backpropagation to strengthen my understanding through coding practice.

    One of the main challenges I’m facing is fully understanding the mathematical intuition behind backpropagation. While I understand the general process, some of the details are still unclear. Additionally, translating theory into code can be difficult at times, especially when debugging model behavior. At the moment, I don’t urgently need instructor assistance, but I may seek help if I run into issues while implementing my full model next week.

    Overall, this week was productive and helped deepen my understanding of how neural networks function internally. I’m starting to feel more confident moving from theory to practice, though I know I still need more hands-on experience to fully master these concepts.

Comments

Post a Comment

Popular posts from this blog

CST 489 - Week 4

 Hey everyone, Since I have chosen to take part in the Certificate Track of the capstone project, I have faced some challenges during this week's assignments. This week, I spent time developing my certificate track capstone proposal, which helped me clarify my learning goals and better understand how each certification connects to my long-term career plans. Writing the proposal forced me to think beyond just completing courses and instead focus on the skills, artifacts, and outcomes I want to gain from each certificate. I also spent a significant amount of time trying to figure out how to enroll in the AWS Cloud Practitioner course. Navigating AWS’s training site was more complicated than I expected, and it took some trial and error to understand the different learning paths and options available. Although it was a bit frustrating, the experience helped me become more comfortable with AWS’s ecosystem and reminded me that problem-solving and persistence are essential parts of learni...
Read more

CST 499 - Week 1

 Hey everyone,      This week, I completed three modules: Tools of the Trade: Linux and SQL, Assets, Threats, and Vulnerabilities, and Sound the Alarm: Detection and Response . Each module built on the last and helped me better understand both the technical and analytical sides of cybersecurity .      In Tools of the Trade: Linux and SQL , I strengthened my understanding of how operating systems and databases are used in security work. I learned how Linux is widely used in cybersecurity for system administration, log analysis, and security operations. Practicing basic Linux commands helped me see how security analysts navigate file systems, manage permissions, and investigate system activity. I also learned how SQL is used to query databases, filter information, and analyze large sets of data. Understanding how to retrieve specific information from logs and datasets showed me how powerful SQL can be in identifying suspicious patterns or investigating ...
Read more

CST 489 - Week 7

 Hey everyone, This week, I expanded my understanding of AWS databases, security, governance, migration, pricing, and architectural best practices. I reviewed database services including Amazon RDS and Aurora for relational workloads, DynamoDB for NoSQL applications, Neptune for graph data, and DocumentDB for document-based use cases. I also learned how Amazon ElastiCache improves performance through in-memory caching. A key takeaway was understanding when to choose managed services versus unmanaged deployments on EC2. I explored migration services such as AWS DMS, AWS Schema Conversion Tool (SCT), Migration Hub, DataSync, and Transfer Family. I reinforced knowledge of the 7 Rs migration strategies and the AWS Cloud Adoption Framework (CAF), particularly the importance of change management through the People perspective. On security and governance, I reviewed IAM best practices (principle of least privilege), AWS KMS for encryption key management, AWS Shield for DDoS protection, AW...
Read more