In this repository you will find some of my Data Science work and technical skills.
At the moment, you can find a machine learning model created with Python (SciKit) which can predict the sentiment of any gaming review, and a statistical analysis tool that can calculate the estimated value of a Magic: The Gathering box.
You can either download the Python script named 'sentiment-ml' (Jupyter Notebook format) and run it to see how the model performs. Feel free to run further optimisations! You might want to download the datasets in the 'data' folder, too.
In case you just want to play with the model itself, you just need the following files:
- count_vect (the vectoriser)
- data/finalized_model.sav (the model)
- loading_model.ipynb (the script to load the model)
You can then run the 'test_reviews.csv' dataset through the model or write your own review. Your choice!
- random
- pandas
- json
- sklearn
- pickle
The sentiment Machine Learning project was created using JupyterLab 3.0.11 and tested on Google Colab. Should you run into any error, please reach out!
Please download the Python script named 'sentiment-ml' (Jupyter Notebook format), together with both datasets (CSV files) in the 'data' folder.
While the box generator function might be the star of the script, it also includes key datapoints via statistical analysis and a few charts to understand the data in a more friendly manner.
Also, if you have data from other MTG sets, you can run this same script to calculate the new set.
- numpy
- pandas
- matplotlib
- datetime
- statistics
The statistical model project was created using JupyterLab 3.0.11 and tested on Google Colab. Should you run into any error, please reach out!