Indego bike project https://www.rideindego.com/about/data/ has made their data available for public to analyze, Indego Bhavya.pdf is collection of small scripts used for data wrangling process.
Face recognition examples in face_reco.ipynb
logistic_regression_from_scratch.ipynb : the notebook provides logistic regression model from scratch with inputs and weights
sigmoid_and_relu_FF.ipynb : python code that takes 5 feature values and computes two targets based on random values for all weights and biases and gives the user the option of choosing either a RELU or a sigmoid function. The code is for feed forward only, for Relu and Sigmoid functions and is from scratch without using any tools like keras.
complete NN.ipynb is the complete neural network i.e feed forward and back propogation, using sigmoid and relu function without using keras. weights and biases are randomly generated.
keras NN.ipynb is nueral network designed for a toy problem using sigmoid and relu function.
diabetes prediction.ipynb: neural network to predict whether or not an individual develops diabetes. used pima-indians-diabetes.csv as input data.
congressional voting predictions.ipynb: neural network to predict whether the politician is republican or democrat based on their voting record. used votingrecords.csv as input data.
overfit_underfit.ipynb: over fitting, under fitting and fit examples for neural network,abalone.data.txt data used for inputs.
dropout NN.ipynb: this notebook implements drop out, model validation with training and testing and then saves the final model
checkpoint.ipynb: this notebook uses checkpointing to save the model weights and biases by taking "snap shots" while the model is running.