Repo of various notebooks I've written to explore concepts, libraries, packages, and everything in-between.
If not written by me, source is attributed.
Implements a neural network from scratch on the MNIST dataset. Also, Implements a few extras like DropConnect, ADAM Optimzation, and Generalization Loss Early Stopping
Basic CNN using the Dogs vs. Cats Dataset
Basic Feed Forward / Multi_Layer Perceptron using Tensorflow and Iris Dataset.
Basic Feed Forward / Multi_Layer Perceptron using Tensorflow and Iris Dataset. Uses Tensorflow API best practices like layers and estimators
Basic CNN using the Dogs vs. Cats Dataset
Basic Multi-Layer Perceptron using Tensorflow and MNIST dataset (Source)
Basic Multi-Layer Perceptron using Tensorflow and MNIST dataset. Uses Tensorflow API best practices like layers and estimators (Source)
Using Frisch–Waugh–Lovell Theorem for Causality Estimation of treatment variables
A notebook that looks are feature importances provided by the skater python package
Bayesian Regression using PyMC3
Explores user-to-user recommendation systems like similarity measures, Matrix Factorization, and how they are leveraged in Deep Learning.
Ensemble stacking example
Linear Regression(s) from scratch
- Linear Regression
- Ridge Regression
- Polynomial Regression
Compares Linear Regression's closed form solution against a gradient descent implementation
Logistic Regression from scratch
Survey of optimization libraries in python. The notebook is mainly focused on hyperparameters, but you can use the libraries for any general optimization problem.
Softmax Regression (Multi-Class Logistic Regression) from scratch
Support Vector Machine from scratch
Various MAB algorithms
Overview of the Beta Distribution
MLE parameter estimation
Intuitions and why you'd use them about highly used distributions
Bayesian Parameter Estimation using pymc3