manvillej / python-andrewngml

Python implementation of Andrew Ng's ML course projects

Python-AndrewNgML

Python implementation of Andrew Ng's ML course projects

• Ex1 (Linear Regression) = Complete!
• Ex2 (Logistic Regression & Regulation) = Complete!
• Ex3 (MultiClass LR and Neural Network Prediction) = Complete!
• Ex4 (Multilayer Neural Network) = Complete!
• Ex5 (Variance vs Bias using Linear Regression) = Complete!
• Ex6 (Support Vector Machines and Email Spam Classification) = Complete!
• Ex7 (Kmeans & PCA for data compression and Visualization) = Complete!
• Ex8 (Anomaly Detection & Recommender Systems) = Complete!

python-andrewngml's Issues

Create file to replicate environment requirements

optimizer throwing error

optimization is throwing error when calling cost function: nnCostFunction in ex4helper.

Problem is when h2 is 0 or 1, the log value throws an error. I am separating the two conditions so a log of a zero value will never happen using the np.extract method.

Add Docstrings to all functions

article on docstrings

PEP 257

Ex8

Need to complete Ex8 - Anomaly Detection & Recommender Systems

Refactor code

Code should be refactored to include a main() and limit line length to <80 characters

Ex6

Need to Complete Ex6 - Support Vector Machines

Rewrite Readme

rewrite readme to be more helpful

Ex4

Need to complete Ex4 - multilayer neural networks

ValueError: iterator is too large when running ex7_pca.py on raspberrypi

ex7_pca throws a ValueError: iterator is too large at line 141,when running the program on a rapsberry pi 3. The current memory limitations and the 32 bit OS makes the operation too expensive for the program.

I tried switching the array type from float64 to float32, but it didn't succeed

Since this issue is limited to more of a problem with the raspberrypi than the code. I am going to just leave it there for now

Ex5

Need to complete Ex5 - regularized linear regression & Bias vs Variance

Ex7

Need to complete Ex7 - K-mean Clustering & Principal Component Analysis