Final Julia project

The project is divided into two parts:

1. Named Entity Recognition using LSTM performed on conll2003 and OntoNotes5.0 datasets
2. Titanic survival classification using KNN and Logistic regression

Project Installation

Package is not officially registered but you can download it using following command:

(@v1.8) pkg> add https://github.com/B0B36JUL-FinalProjects-2022/Projekt_lamossta

NER structure

• src/nn.jl => implemented classification LSTM net
• src/prepare_ner_data.jl => dataset loading, embedding conversion and feature vector construction

Titanic structure

• src/knn.jl => implementation of KNN classification algorithm
• src/logreg.jl => implementation of Logistic regression
• src/prepare_titanic_data.jl => CSV loading, missing data augmentation and feature vector construction

Examples and tests

• examples/example.jl => example classification using all implemented algorithms mentioned above
• examples/ner_ntb.ipynb => notebook with showcase of NER classification
• examples/titanic_ntb.ipynb => notebook with showcase of Titanic classification
• tests/runtests.jl => Unit tests for various functions