ML-Text-Classification

Machine Learning Algorithm Toolbox

• Different shallow and deep learning algorithms for text classification

Proposed Directory Structure

data
|--- classif_data.csv/tsv/txt
src
|--- DataLoader.py
|--- Models.py
|--- Trainer.py
|--- Inference.py
|--- FeatureExtractor.py
main.py

Expalination

Central to any ML system are three key things:

1. data, on which model will be trained;
2. features, the representation of data that will be the input to the model; and
3. algorithm (or model itself), which is going to be trained

A simple pipeline of any ML project can be defined as:

1. Prepare your data - split them into train and test sets. We'll do this using DataLoader.py
2. Represent your data - extract features or embed your data, can also be considered the pre-processing step. We'll do this usinf Extractor.py
3. Train the model. Will be done in Trainer.py
4. Predict using the model. Will be done using Inference.py

main.py will be a high-level wrapper to call different classes at a single place.