/Coursera-Machine-Learning-Foundations

Primary LanguageJupyter NotebookMIT LicenseMIT

Machine Learning Foundations

Code for programming assignments in Python from the Coursera's course Machine Learning Foundations, taught by Carlos Guestrin and Emily Fox both of them from Washington University.

Week 1 - Introduction (No assignment)

  • getting-started-with-iPython-Notebook.ipynb - Introduction to iPython notebook.
  • getting-started-with-SFrames.ipynb - Introduction to SFrame.
  • people-example.csv - Input data for week 1.

Week 2 - Regression: Predicting House Prices (100%)

  • predicting-house-prices.ipynb - iPython notebook with a regression model developed for prediction of house prices.
  • home_data.gl - Database containing information about houses in the US (size, number of bathrooms, number of rooms, and so on) that was used to develop the aforementioned regression model.

Week 3 - Classification: Analyzing Product Sentiment (100%)

  • analyzing-product-sentiment.ipynb - iPython notebook with a classification model developed for recommending products on Amazon.
  • amazon_baby.gl - Amazon database containing nformation about products ratings and reviews hat was used to develop the aforementioned classification model.

Week 4 - Clustering and Similarity: Document Retrieval (100%)

  • document-retrieval.ipynb - iPython notebook with a clustering model developed to find people releated to a given input.
  • people_wiki.gl - Wikipedia database containing articles about famous people.

Week 5 - Recomending Products: Song Recommender (100%)

  • song-recommender.ipynb - iPython notebook with a recommender model developed to find similar songs or recommend songs to a given user.
  • song_data.gl - Database containing information about musical preferences of users.

Week 6 - Deep Learning: Deep Features for Image Classification and Retrieval (100%)

  • deep-features-for-image-classification.ipynb - iPython notebook with a neural network trained to find similar images and clasify them in into one of four categories. Moreover, use transfer learning from the neural network ImageNet to compare perfomance against a neural network trained with millions of images.
  • deep-features-for-image-retrieval.ipynb - iPython notebook with a neural network developed to find similar images in the dataset.
  • image_test_data.gl - Folder with database for traininput data for week 2.
  • image_train_data.gl - Folder with input data for week 2.