Ngugi Cyrus

Electrical engineer
Music
Analyst

[Ngugi Cyrus @Cyrusz55]

Music is great manze

[Ngugi Cyrus @Cyrusz55]

@_falsi1ke @SafaricomPLC

[ROYZ @royzkingin]

To achieve anything significantly big in life, you must be disgusted by mediocrity.

[Ngugi Cyrus @Cyrusz55]

@VkBoit Nakuja kukutafuta. Hiyo snack lazima ununue😆

[Malvin @ManOfFocus_]

REAL TALK: Marrying the right woman is the most important part of building wealth.

[Ngugi Cyrus @Cyrusz55]

@RutoJames7113 What I am saying, being religious does not prevent them from learning science

[Ngugi Cyrus @Cyrusz55]

@Complexleol @uncleparis30 @magattew It would be better if you share your knowledge.

[Ngugi Cyrus @Cyrusz55]

@Kamau266 @x_weeep @kigothor Do you have solid proof? 80% inakaa guesswork

[Ngugi Cyrus @Cyrusz55]

@KevvinMwangi @Kamau266 What about 2 years after that

[Ngugi Cyrus @Cyrusz55]

Perhaps there exist an alternate universe where human programmers are better than ai. Literally.
The sky is the limit

[Ngugi Cyrus @Cyrusz55]

@LuxDevHQ
In my attempt to learn how to deploy models and move from my localhost, I pushed my model to huggingface and hosted my frontend and backend on onrender. Check out my salary prediction site. The model's test r2 score is 0.63
https://t.co/P6MVRYA44t
#DataScience #ML

[Ngugi Cyrus @Cyrusz55]

@LuxDevHQ
In my attempt to learn how to deploy models and move from my localhost, I pushed my model to huggingface and hosted my frontend and backend on onrender. Check out my salary prediction site. The model's test r2 score is 0.63
https://t.co/P6MVRYA44t
#DataScience #ML

[Ngugi Cyrus @Cyrusz55]

here are the scores.

[Ngugi Cyrus @Cyrusz55]

Built an end-to-end Japan used car price predictor: scraping, feature engineering, model training, FastAPI, and a polished frontend. https://t.co/eIqMCdf4d2

#MachineLearning #FastAPI #DataScience

[Ngugi Cyrus @Cyrusz55]

Make it I will. At all cost!!!

[Ngugi Cyrus @Cyrusz55]

@_mwangangikyalo Mzee😂😂

[Ngugi Cyrus @Cyrusz55]

@akinyi__wendy @unclebourne Please answer the question above. I am also very curious

[Ngugi Cyrus @Cyrusz55]

I would appreciate any suggestions, criticizm, or any other helpful info

[Ngugi Cyrus @Cyrusz55]

🚀
Built a FastAPI interface powered by XGBoost to predict annual salaries using real Stack Overflow survey data. The model analyzes 10 key factors including experience, education, location, and employment type.
this was a task given by @ luxdev_HQ
https://t.co/TSmCDhzwxM

[Ngugi Cyrus @Cyrusz55]

🔧stack:

✓ XGBoost regressor
✓ FastAPI backend
✓ Vanilla HTML/CSS/JS frontend
✓ Response time: 182ms average
Model Performance:
✓ CV R²: 0.4949 (explains ~49% of salary variance)
✓ Test RMSE: 0.96 (log-space = ~10-15% error on actual salaries)