Advanced Topics in Analytics

Instructor: Sergio A. Mora Pardo

• email: sergioa.mora@javeriana.edu.co
• github: sergiomora03

Knowledge of the challenges and solutions present in specific situations of organizations that require advanced and special handling of information, such as text mining, process mining, data flow mining (stream data mining) and social network analysis. This module on Natural Language Processing will explain how to build systems that learn and adapt using real-world applications. Some of the topics to be covered include text preprocessing, text representation, modeling of common NLP problems such as sentiment analysis, similarity, recurrent models, word embeddings, introduction to lenguage generative models. The course will be project-oriented, with emphasis placed on writing software implementations of learning algorithms applied to real-world problems, in particular, language processing, sentiment detection, among others.

Requiriments

• Python version >= 3.7;
• Numpy, the core numerical extensions for linear algebra and multidimensional arrays;
• Scipy, additional libraries for scientific programming;
• Matplotlib, excellent plotting and graphing libraries;
• IPython, with the additional libraries required for the notebook interface.
• Pandas, Python version of R dataframe
• Seaborn, used mainly for plot styling
• scikit-learn, Machine learning library!

A good, easy to install option that supports Mac, Windows, and Linux, and that has all of these packages (and much more) is the Anaconda.

GIT!! Unfortunatelly out of the scope of this class, but please take a look at these tutorials

Evaluation

• 50% Project
• 40% Exercises
• 10% Class participation

Deadlines

Session	Activity	Deadline	Comments
Deep Learning	Exercises Project	March 21th	Expo March 22th
NLP	Exercises Project	April 25th April 11th	Expo April 12th
Graph Learning	Exercises Project	May 24th
Final grade	project	May 31thth

Slack Channel

Join here!

Schedule

Basic Methods MLOps

Date	Session	Notebooks/Presentations	Exercises
March 1st	Machine Learning Operations (MLOps)	Intro MLOps
March 1st	ML monitoring & Data Drift	Intro Data Drift L2 - Intro Data Drift L3 - Intro Model Monitoring	E1 - Data Drift in Used Vehicle Price Prediction
March 1st	Machine Learning as a Service (AIaaS)	1 - Intro to APIs L1 - Model Deployment	E2 - Model Deployment in Used Vehicle Price Prediction

Intro Deep Learning

Date	Session	Notebooks/Presentations	Exercises
March 8th	First steps in deep learning	3 - Intro Deep Learning L3 - Introduction Deep Learining MLP L4 - Simple Neural Network (handcraft) L5 - Simple Neural Network (Images) L6 - Deep Learning with Keras L7 - Deep Learning with Pytorch	E3 - Neural Networks in Keras and PyTorch
March 15th	Deep Computer Vision	4 - Convolutional Neural Networks L5 - CNN with TensorFlow L6 - CNN with PyTorch 🔥 L7 - Tranfer Learning with TensorFlow	E4 - Tranfer Learning with PyTorch
March 22th	Computer Vision Project	Exercises Deadline	P1 - Frailejon Detection (a.k.a "Big Monks Detection")

Intro Natural Language Processing

Date	Session	Notebooks/Presentations	Exercises
March 22th	Introduction to NLP	1 - Introduction to NLP 2 - NLP Pipeline E1 - Tokenization

Text Representation

Date	Session	Notebooks/Presentations	Exercises
March 22th	Space Vector Models	1 - Basic Vectorizarion Approaches L2 - OneHot Encoding L3 - Bag of Words L4 - N-grams L5 - TF-IDF L6 - Basic Vectorization Approaches	E2 - Sentiment Analysis
March 29th	Distributed Representations	2 - Word Embbedings L7 - Text Similarity L8 - Exploring Word Embeddings L9 - Song Embeddings L10 - Visualizing Embeddings	E2 - Homework Analysis (Bonus) E3 - Song Embedding Visualization E4 - Spam Classification

NLP with Deep Learning

Date	Session	Notebooks/Presentations	Exercises
April 5th	Deep Learning in NLP (RNN, LSTM, GRU)	4 - RNN, LSTM, GRU L12 - NLP with Keras L11 - NLP with Keras L13 - Recurrent Neural Network and LSTM L14 - Headline Generator	E5 - Neural Networks in Keras for NLP E6 - Neural Networks in PyTorch for NLP E7 - RNN, LSTM, GRU
April 12th	NLP Project		P1 - Movie Genre Prediction
April 12th	Attention, Tranformers and BERT	5 - Encoder-Decoder 6 - Attention Mechanisms and Transformers 7 - BERT and Family L16 - Positional Encoding L17 - BERT for Sentiment Clasification L18 - Transformers Introduction	E8 - Text Summary E9 - Question Answering E10 - Open AI
April 19th	Holy Week	Holy Week	Holy Week
April 25th		Exercises Deadline

Intro Graph

Date	Session	Notebooks/Presentations	Exercises
April 26th	Intro to Graphs	Intro to Graphs L19 - Intro to Graphs
April 26th	Graphs Metrics	L20 - Graph Metrics L21 - Graphs Benchmarks L22 - Facebook Analysis	E10 - Twitter Analysis

Graph Representation Learning

Date	Session	Notebooks/Presentations	Exercises
May 3th	Graph Representation	Graph Representations L23 - Graph Embedding L24 - Deep Walk L25 - Node2Vec L26 - Recommendation System with Node2Vec	E11 - Patent Citation Network (Node2Vec with RecSys)

Intro to Geometric Deep Learning

Date	Session	Notebooks/Presentations	Exercises
May 10th	Graph Neural Network	L27 - Pytorch intro with Node2Vec L29 - Graph Neural Networks - Node Features L29 - Graph Neural Networks - Adjacency Matrix L31 - Graph Convolutional Networks - Node Classification L33 - Graph Convolutional Networks - Node Regression L34 - Graph Attention Networks	L30 - Graph Neural Networks - Facebook Page-Page dataset L32 - Graph Convolutional Networks - Facebook Page-Page dataset L34 - Graph Attention Networks - Cite Seer
May 17th	Graph Machine Learning Task [Optional]	L35 - Graph AutoEncoder - Link Prediction L36 - Graph Variational AutoEncoder - Link Prediction [extra] L37 - Node2Vec - Link Classification L38 - Graph Isomorphism Network - Graph Classification
May 24th	Geometric Deep Learning Project	Exercises Deadline	P3 - Graph Machine Learning / P3 - Graph Machine Learning [old < 2022]

Grades

Date	Session	Notebooks/Presentations	Exercises
May 31th	Final Grades

Interest Links 🔗

Module	Topic	Material
NLP	Word Embedding Projector	Tensorflow Embeddings Projector
NLP	Time Series with LSTM	ARIMA-SARIMA-Prophet-LSTM
NLP	Stanford	Natural Language Processing with Deep Learning
GML	Stanford	CS224W: Machine Learning with Graphs

Extra Material

Module	Topic	Material
NLP	Polarity	Sentiment Analysis - Polarity
NLP	Image & Text	Image Captions
ML	Hyperparameter Tuning [WIP]	Exhaustive Grid Search Randomized Parameter Optimization Automate Hyperparameter Search
NLP	Neural Style Transfer	Style Transfer