MED07 / FSR

Course Content

Introduction

Machine Learning

1. Supervised Learning,
2. Unsupervised Learning

Natural Language Processing

1. Regular Expressions,
2. Tokenization,
3. Character Encoding,
4. Part-of-Speech Tagging,
5. Chunking,
6. Stemming and Lemmatization,
7. Parsing,
8. Named Entity Recognition,
9. Topic Segmentation

Introduction to Deep Learning for NLP

1. Sequence models,
2. Embeddings,
3. BERT models

Practical Lab setup

You need a google account for using Colab for the first labs. Later, you have to use a virtual environment like conda to run the jupyter notebooks locally.

On local: Install Miniconda

Download

Miniconda from https://conda.io/miniconda.html

install

$bash Miniconda3-latest-Linux-x86_64.sh

Create environment

Create env

$conda create --name nlplab python=3.6

activate env

$source activate nlplab

Install packages

(nlplab)$pip3 install --upgrade pip

(nlplab)$pip3 install numpy scipy pandas matplotlib

(nlplab)$pip3 install scikit-learn

(nlplab)$pip3 install jupyter

(nlplab)$pip install --ignore-installed --upgrade tensorflow

list packages

(nlplab)$ conda list

Jupyter notebook

run jupyter

(nlplab)$ jupyter notebook

install package inside a jupyter notebook cell

!pip install numpy