https://github.com/salihcanaydogdu/nlp_medicaltranscriptions

NLP Project for Medical Transcriptions.For more details, you can read my report and read.me
https://github.com/salihcanaydogdu/nlp_medicaltranscriptions

artifical-intelligense deep-learning health-data natural-language-processing

Last synced: 28 days ago
JSON representation

NLP Project for Medical Transcriptions.For more details, you can read my report and read.me

• Host: GitHub
• URL: https://github.com/salihcanaydogdu/nlp_medicaltranscriptions
• Owner: SalihCanAydogdu
• Created: 2024-09-19T05:32:14.000Z (4 months ago)
• Default Branch: main
• Last Pushed: 2024-09-19T07:17:24.000Z (4 months ago)
• Last Synced: 2024-11-25T05:22:44.727Z (about 1 month ago)
• Topics: artifical-intelligense, deep-learning, health-data, natural-language-processing
• Language: Jupyter Notebook
• Homepage:
• Size: 8.77 MB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
Project Purpose: This project aims to classify medical specialties through medical transcripts.

Dataset: Medical transcript dataset from Kaggle was used. (Kaggle link: https://www.kaggle.com/tboyle10/medicaltranscriptions)

Preprocessing: Noise reduction, data cleaning and lemmatization processes were applied.

Text Vectorization: Texts were converted to vectors using Bag-of-Words and TF-IDF methods. Dimensionality reduction was done with PCA.

Algorithms: Multinomial Naïve Bayes, Random Forest, Xgboost, LightGBM and CNN + LSTM (Ensemble Learning) algorithms were applied.

Conclusion: Xgboost and Ensemble Learning methods gave the most successful results. The project can be improved by strengthening it with more data.