Introduction to Data Science @ FGV

Instructor: Renato Rocha Souza

This is the repository of code for the "Introduction to Data Science"

This class is about the Data Science process, in which we seek to gain useful predictions and insights from data. Through real-world examples and code snippets, we introduce methods for:

• data munging, scraping, sampling andcleaning in order to get an informative, manageable data set;
• data storage and management in order to be able to access data (even if big data);
• exploratory data analysis (EDA) to generate hypotheses and intuition about the data;
• prediction based on statistical learning tools;
• communication of results through visualization, stories, and interpretable summaries

Detailed Syllabus:

• Propaedeutic ref1, ref2, ref3, ref4, ref5

• Related Courses cs109, cs229, ML Andrew Ng

• Data Science Concepts ref1, ref2, ref3, ref4, book1, book2, book3

  • Data Scientists ref1, re2, ref3, learning path
  • Statistics book1, book2, ref1, ref2, ref3
  • AI ref1
  • Machine Learning ref1, ref2
  • Business Intelligence ref1

• Data Science Process ref1, ref2

  • Data Formats ref1
  • Data Acquisition ref1
  • Data Engineering ref1, ref2, ref3
    • Handling Missing Values ref1
    • Changing Datatypes ref1
  • Exploratory Data Analysis ref1, book
  • Data Visualization ref1, ref2

• Model Selection ref1

  • Feature Engineering ref1, book
    • Automated Feature Engineering
      featuretools
  • Feature Selection ref1
    • Numeric Data ref1
    • Discrete/Categorical Data ref1
    • Textual Data ref1, ref2
  • Hiperparameter Search ref1
  • Cross Validation ref1, video1
  • Oversampling and Undersampling ref1
  • Regularization ref1, ref2
  • Bias and Variance ref1
  • Overfitting and Underfitting ref1
  • Evaluation Metrics ref1, ref2

• Machine Learning Algorithms ref1, ref2, ref3, ref4, ref5, ref6

  • Unsupervised ref1

    • Dimensionality reduction ref1
      • PCA ref1
      • SVD ref1
      • t-SNE ref1
    • Clustering ref1, ref2, ref3
      • K-Means ref1
      • Hierarchical Clustering ref1
      • K-Modes ref1
    • Topic Modeling ref1, ref2, ref3, ref4
      • LDA ref1, ref2
    • Unsupervised Deep Learning ref1

  • Supervised

    • Regression ref1, ref2, ref3, ref4, ref5

    • Classification ref1, ref2, ref3, ref3

    • Linear Models ref1

      • Linear Regression ref1

        • Polynomial Regression ref1 ref2
        • Stepwise Regression ref1
        • Ridge Regression ref1
        • Lasso Regression ref1, package
        • ElasticNet Regression ref1

      • Logistic Regression (Categorical) ref1, ref2, ref3, ref4

      • Support Vector Machines ref1, ref2, ref3, ref4, paper, video

        • SVR ref1, ref2
        • SVC ref1, ref2

      • Passive-Aggressive ref1

      • Perceptron ref1, ref2

    • Bayesian Models

      • Naive Bayes ref1
      • Gaussian Naive Bayes ref1
      • Bernoulli Naive Bayes ref1

    • Guassian Mixtures ref1 package

    • k Nearest Neighbors (kNN) ref1

    • Decision Tree ref1, ref2

    • Ensemble Models ref1, ref2 ref3, package

      • Bagging ref1
        • Random Forest ref1
        • Extremely Randomized Trees ref1
        • Bagging Classifier ref1, package
      • Boosting ref1
        • Adaboost ref1
        • GBM ref1
        • XGBoost ref1, ref2, package
        • LightGBM ref1
        • CatBoost ref1
        • Regularized Greedy Forests ref1, package
      • Stacking ref1
      • Voting ref1

    • Genetic Algorithms ref1, ref2

      • TPOT,
      • Auto SKLearn
      • Auto Keras

    • Neural Networks and Deep Learning ref1, ref2, ref3, ref4, ref5, ref6, ref7, book, video, meme

      • Convolutional Neural Networks ref1, ref2, ref3, ref4, ref5, ref6
        • CNNs for NLP
        • CNNs for object detection
        • CNNs for Time Series
      • RNNs (Sequence Models) ref1
        • Word Embeedings ref1, ref2
        • Word2vec ref1, ref2, ref3, en-us trained models, pt-br trained models, ge-de trained models
        • Doc2vec ref1
        • Char2vec ref1
        • Glove ref1
      • LSTMs and GRUs ref1
      • Reinforcement Learning ref1, ref2, ref3, ref4, ref5, ref6, programming resource
      • Transfer Learning ref1, ref2
      • Generative Adversarial Networks ref1, ref2, ref3
      • Neural Network concepts

        • Weight Initialization ref1, ref2

        • Weight Averaging ref1

        • Hyperparameter Tuning ref1, ref2

        • Gradient Descent ref1, video

        • Backpropagation ref1

        • Neural Network Visualization

• Data Science Tasks

  • Time Series Analysis ref1, ref2, ref3, ref4, video
  • NLP and Text Mining ref1, ref2
  • Information Retrieval
  • Graphs and Network Analysis
  • Sentiment Analysis ref1, ref2, ref3, ref4, ref5, ref6, ref7
  • Recommender Systems ref1, ref2, ref3, ref4

• Preparing the Environment

  • Linux, Python, Virtualenv

• Data Science and Visualization Tools

  • Versioning Tools
    • Git
    • Github
    • GIT Large File Versioning
    • Gitlab
  • Exploratory Data Analysis Tools
    • Jupyter ref1, ref2, ref3, ref4, ref5
    • Numpy ref1, ref2
    • Pandas ref1, ref2, ref3
      • Pandas Machine Learning
      • SKLearn Pandas
      • Geopandas
      • Framequery
    • Statsmodels
  • Machine Learning Tools
    • Scikit-Learn
      • Inbalanced Learn
      • ForestCI
    • Tensor Flow ref1, ref2
    • Keras ref1, ref2
    • Gensim
    • Orange
  • NLP Tools
    • NLTK
    • Spacy
    • TextBlob
  • Visualization Tools
    • Matplotlib
    • Seaborn
    • Bokeh ref1, example
    • Plotly
    • Altair
    • GGPlot2
    • MPLD3
      • d3.js
    • HoloViews
    • Folium
      • Leaflet
    • Neural Networks visualization
      • TensorBoard
      • Tensor Flow Playground
      • Deep Replay
  • Big Data and Distributed computing
    • Map Reduce
    • Spark
    • Dask
  • Analytical Pipelines
    • Scikit-Learn Pipelines ref1
    • Luigi ref1
    • Airflow ref1, ref2
  • Other Tools
    • NetworkX
    • ETE Toolkit
    • ODO
  • Relational databases and SQL
  • NoSQL Databases
    • Elastic Search
    • Graph Databases
      • Graph Theory
  • Machine Learning Datasets ref1

We are using https://git-lfs.github.com because the /datasets files can be large. Install it before the git clone.