s013mmb / notebooks Goto Github PK

Includes material (code + description) that I turned into notebooks, in connection to our AI and machine learning certifications (see details here). You are free to download and run them in your environment.

In particular:

• Interpolation_temperatures_india.ipynb (geospatial statistics) is a simplified solution, part of project 3.2 in the project textbook.
• GAN_diabetes.ipynb (generative adversarial network, diabetes dataset) is part of project 5.2 in the project textbook.
• sd_vendors (GAN vendor comparison, evaluation metric, holdout method) is part of project 5.1 in the project textbook.
• NoGAN.py (very fast better than GAN synthesizer for tabular data, without neural networks) is part of project 2.1 in the project textbook.
• Copula_insurance_byGroup.ipynb and Copula_insurance_nogroup.ipynb (insurance dataset, tabular data generation with copulas) are related to project 5.2.
• DeepResampling.ipynb is a different version of NoGAN. The full name of the method is Hierarchical Deep Resampling (no neural network involved) or NoGAN2. Also very fast and better than GAN, with self-tuning, and explainable AI.

The project textbook is offered to participants in the certification programs. For details about the cost (starting at $44) and how to enroll, see here.