Skip to content

A repositiory of the projects I worked on or currently working on. The projects are written in Python (Jupyter Notebook). Feel free to click on the projects to see full analysis and code.

Notifications You must be signed in to change notification settings

Dancar96/Data_Science_Portfolio

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

17 Commits
 
 
 
 

Repository files navigation

Data_Science_Portfolio

A repositiory of the projects I worked on or currently working on. The projects are written in Python (Jupyter Notebook). Feel free to click on the projects to see full analysis and code.

  • First programming project with Python in partnership with https://github.com/laura-pc of the famous classic game 'Battleship'.
  • This project has been realized using the NumPy Python library. The boards are generated in the terminal.
  • First EDA trying to discover with data who is the greatests singer or music band of all time by Daniel Carrera
  • This project has been realized using Python and the libraries: NumPy, Pandas, MatplotLib, Seaborn and Plotly. It has also been used Selenium and Beautiful Soup 4 for Web Scraping.
  • By analising the music industry sales trend and analyzing by sales, awards, gold and platinum records, Grammys, charts, Spotify plays, hits and tenure on The Hot 100 Billboard and Rolling Stones magazine rankings, an analysis has been made to end the debate of who is the best, most influencial musician of all time.

  • A ML project for unmask non-legitimate URL's.
  • This first project has been realized using Python and the libraries: NumPy, Pandas, MatplotLib and Scikit-Learn.
  • For the choice of the ML model, a comparison of supervised language models was made according to their Recall evaluation metric. It has been decided to use this metric since it is the most optimal for our data, given that if we are going to classify URL's as legitimate or phishing, we must prioritize that no phishing URL is classified as legitimate, it is preferable that if the prediction is wrong it is because it classifies a legitimate URL as phishing than the other way around.
  • After the comparison of models according to their recall score, the model chosen is 'XGBClassifier' with the optimal hyperparameters for the dataset according to the 'GridSearchCV' performed.
  • The model is learned and then saved as a pickle file.

Corr_Martrix Model_Comparison

About

A repositiory of the projects I worked on or currently working on. The projects are written in Python (Jupyter Notebook). Feel free to click on the projects to see full analysis and code.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published