NLP

• This repository explores Natural Language Processing (NLP) in great detail. Further details are provided below, .ipynb uses python package.

Folder	Description	Libraries
0. Expl Data Analysis	Exploratory Data Analysis on Text data WordCloud, Missing Value, HeatMap, Box plot etc	Seaborn, Matplotlib, Pandas, pandas_profiling
1. Text Classification	Text Classification using Rule based and Machine learning based approch like Naive Bayes, Bag-of Words approche	scikit learn, VADER, Textblob
2. Text PreProcessing	Stemming, Lemmatization,Part of Speech Tagging, spelling correction, dealing with punctuataion,stopwords, non alphabet, lowercase etc.	Textblob, regexp, NLTK
3. Word Embedding	ELMo, FastText, Glove, Word2Vec	glove_python, gensim, tensorflow_hub
4. Keras	Creating Model CNN, RNN, MLP, CuDNNGRU, GRU, Embedding layer, Padding , Truncating	keras, tensorflow
5. mlMastery	preprocessing and Vectorizing Keras , POS Tagger, pretrained word embedding, RNN model fot text classifcation,	Stanford core NLP, keras, tesnorflow,
6. Optimization techq	Cross validation, Starified dataset, Repoducability epochs, optimizer, learning rate, dropout, momentum, grid search	keras
7. Experiments for deployment	pickling, deployment pipeline, model wights, performance evaluation	keras, tensorflow, NLTK
8. Spacy	POS tagging, NER, Collocation, Entity Extraction, contraction expansion	spacy
9. BERT	training BERT classifier (1st offical relased version)	shell script, tensorflow
CNN Img classifcation.ipynb	CNN mobilenet architecture for Img Classifcation	Keas, tensorflow
TFLiteConverter.ipynb	Converting .h5 file to .tflite	tensorflow.contrib, tensorflow
ULMfit & Transformer Network.ipynb	ULMFit paper analysis & Transformer Architecure	NA

Usage

Accompanying data files and additional utility scripts have been left out. The goal is rather to demonstrate experience and achievements.