Udacity_connect

This repo contains a collection of Jupyter Notebooks to accompany the Udacity Connect Intensive Machine Learning Nanodegree. The code is written for Python 2.7, but should be (mostly) compatible with Python 3.

List of Contents

• Week 1 (wk1/)

  • PythonPractice_1.ipynb: Introduction to Jupyter Notebook and basic Python programming (including data types, if and while loops, list comprehension, lambda expression, etc.).
  • PythonPractice_2.ipynb: Introduction to Numpy, including how to create Numpy Array, built-in methods in Numpy array, array indexing/selection/slicing, broadcasting.
  • PythonPractice_3.ipynb: Introduction to Pandas. Topics include inputting data into DataFrame and getting summary information, selection and indexing, conditional selection with DataFrame, etc.
  • PythonPractice_4.ipynb: Introduction to data visualization with Matplotlib and Seaborn.
  • data/: containing one sample dataset for the notebooks and one for exercise.

• Week 2 (wk2/)

  • SklearnTutorial.ipynb: Introduction to scikit-learn (sklearn) and a step-by-step guide of building a machine learning model with sklearn with the Titanic Survival dataset from Kaggle
  • SklearnTutorial-solution.ipynb: The solution to the SklearnTutorial.ipynb notebook.

• Week 3 (wk3/)

  • RegressionModels.ipynb: Introduction to the implementation and evaluatoin of regression models to predict housing price with sklearn.
  • RegressionModels-solution.ipynb: The solution to the RegressionModels.ipynb

• Week 4 (wk4/)

  • NeuralNets_Miniproject.ipynb: Introduction to the fundamentals of neural networks, the implementation of single layer and multi layer perceptrons, and perceptron with scikit-learn.
  • NeuralNets_Miniproject-solution.ipynb: The solution to the NeuralNets_Miniproject.ipynb notebook.

• Week 5 (wk5/)

  • BayesNLP_Miniproject.ipynb: Introduction to Bayes theorem and application of Bayes theorem in natual language processing. Write Python methods to calculate maximum likelihood of a word based on the preceding word, and build a Bayes classifier that computes with a context the optimal label for a second missing word based on the possible words that could be in the first blank.
  • BayesNLP_Miniproject-solution.ipynb: The solution to the BayesNLP_Miniproject.ipynb notebook.
  • Quiz.pdf: some quiz on supervised learning.

• Week 6 (wk6/)

  • Clustering.ipynb: Perform K-Means clustering on the Enron dataset. Visualize different clusters that form before and after feature scaling. Plot decision boundaries that arise from K-Means clustering using two of the features.
  • Clustering-solution.ipynb: The solution to the Clustering.ipynb notebook.

• Week 7 (wk7/)

  • PCA.ipynb: Perform Principal Component Analysis (PCA) on a large set of features to explain as much of the variance as possible in the data using a smaller set of features. Visualize the eigenfaces (orthonormal basis of components) that result from PCA. The dataset omes from "Labeled Faces in the Wild" (LFW), a database of more than 13,000 face photographs designed for studying the problem of unconstrained face recognition.
  • PCA-solution.ipynb: The solution to the PCA.ipynb notebook.

• Week 8 (wk8/)

  • FeatureSelection.ipynb: Introduction to Chi-square test statistics and Pearson's Chi-square test. Learn how to perform univariate feature selection using the 'SelectKBest' class from scikit-learn. Learn how to do recursive feature elimination using the RFE class and RFE with cross-validation (RFECV) from scikit-learn.
  • FeatureSelection-solution.ipynb: The solution to the FeatureSelection.ipynb notebook.

• Week 9 (wk9/)

  • Kaggle.ipynb: Introduction to the process of creating machine learning solutions to a real world problem. Practice kills such as data analysis and visualization, model building, evaluation and optimization on a real world dataset.

• Week 10 (wk10/)

  • MNIST_Demo.ipynb: Introduction to TensorFlow and Keras. Learn how to build neural networks using TensorFlow and Keras to solve a multiclass classification problem. Compare TensorFlow and Keras.