NLP with various methodologies using consumer_complaints kaggle dataset url: https://www.kaggle.com/datasets/kaggle/us-consumer-finance-complaints
00: EDA with pandas profiling report, data munging, and feature engineering.
00a: Data preparation for validation dataset.
00b: Data preparation for test dataset.
01: NLP with TfidfVectorizer from sklearn, product classification with several classification models (including feature-engineered variables from EDA), hyperparameter optimization with Optuna, and final ensemble model.
02: NLP with Gensim's Doc2Vec embedded vectors and product classification with several classification models.
03: NLP with simpleTransformers multi-classification model and GPU acceleration. LLMs evaluated: bert-base-uncased and bert-large-uncased.