🚀 Exciting Learning Opportunity: Machine Learning Design Patterns and Principles with Reza Ardestani 🚀

We're thrilled to announce an upcoming session that provides the toolkit and technics helping researchers and practitioners to conduct their work with remarkable ease.

In today's fast-paced world, agile and rapid development is the key, encouraging developers to use code auto generation tools like ChatGPT. These tools, however, are not able to tackle large-scale real-world ML projects, unless you have a command over the different concepts, patterns, and the big picture of your ML projects. This session is aiming to provide these concepts some of which are listed as follows:

Data Handling: Batching and padding, Masking, imbalanced dataset

Model Training: Loss function, Loss function, Time efficiency, Vectorization, Penalization

Optimization and Resource Management: Caching, Profiling, Hyper-parameter organization, Memory management, Computational resources, SLURM commands

Miscellaneous: GitHub for documentation, Structuring a machine learning project/thesis proposal, etc

"By leveraging these concepts, I proposed a model that outperformed counterparts developed by Microsoft and Facebook. Additionally, it expedited the development process by three months as well." Reza said.

We are looking forward to seeing all of you.