Roadmap

• Python Crash Course

  • Intro to python syntax
  • Intro to Pandas/NumPy/MatPlotLib

• Linear regression [supervised]

  ** Supply them with the data, function headers, and plotting/analysis completed. They'll just write the algo.

  • Simple
    • Python: Pandas, NumPy
  • Multivariate
    • Linear algebra: Matrices, matrix product, matrix transpose
    • Stats: Variance, (expected value??)
    • Python: Pandas, NumPy

• Naive Bayes Classifier [supervised]

  ** Supply them with the data, function headers, and plotting/analysis completed. They'll just write the algo.

  • MNIST B/W
    • Binary hypothesis testing
    • Linear Algebra: Matrices, matrix product
    • Stats: Cost functions, probability/expected value
    • Python: Pandas, NumPy

• K-Nearest Neighbors Classifier [supervised]

• Data Visualizations

  ** Walk through a story of how we created visualizations on above 3 supervised learning workshops

• K-Means Clustering [unsupervised] https://archive.ics.uci.edu/ml/datasets/Sentiment+Labelled+Sentences Use kmeans to cluster data of varying length and then attempt to classify it as positive or negative with an SVM. 3 part series. 1.) Data cleansing with Pandas 2.) Preliminary visualizations 3.) Clustering 4.) Visualize clusters 5.) SVM classifier 6.) Review results

• Hopfield Network [memory]

Other ideas

• AWS Sagemaker?
• OpenCV?
• Education datasets (CS 125, 225, 233)?
  • Data cleansing with Pandas
  • Struggling students ML model
  • Visualizations with MPL/Plot.ly