https://github.com/vpereira01/gostart
Playground of Go, ml and prime factorization
https://github.com/vpereira01/gostart
factorization golang jupyter-notebook machine-learning primes
Last synced: 15 days ago
JSON representation
Playground of Go, ml and prime factorization
- Host: GitHub
- URL: https://github.com/vpereira01/gostart
- Owner: vpereira01
- Created: 2022-02-01T19:37:52.000Z (over 4 years ago)
- Default Branch: master
- Last Pushed: 2022-02-02T18:51:35.000Z (over 4 years ago)
- Last Synced: 2025-03-26T09:28:22.495Z (about 1 year ago)
- Topics: factorization, golang, jupyter-notebook, machine-learning, primes
- Language: Jupyter Notebook
- Homepage:
- Size: 81.1 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# What is this?
My playground to learn Go, machine learning and prime factorization.
# Outcomes
* Learned some Go
* Go routines, channels and waitGroups makes it easy to implement multi-threaded workloads
* Performance tuning with `go test -cpuprofile cpu.prof` and `go tool pprof -http localhost:19123 cpu.prof` is also very easy
* Learned some Machine Learning
* Focused on scikit-learn which seems the most easier tool to start on
* Normalization of data with scaling to make raw data processable by ordinary python tools
* That's always good to start by evaluating features/labels correlation with 'seaborn.pairplot'
* Jupyter notebooks are very useful for quick iteration
* Learned about semi-primes factorization
* Initially approach to use machine learning to estimate prime factors sum is not useful:
* Although neural-network prediction score can be high, greater than 0.99, after de-normalization the actual error is very high
* Trying to find the real prime factors sum based on neural-network prediction is way slower than using other methods
* About the Fermat's factorization method
* Tried a different approach, using only algorithms, to find the prime factor difference with a good performance
* Afterwards re-checked Wikipedia and found that I just implemented Fermat's factorization :/