How to train your Deep Neural Network

List of commonly used practices for efficient training of Deep Neural Networks.

Dropout with Theano

Implementing a Dropout Layer with Numpy and Theano along with all the caveats and tweaks.

L1 vs. L2 Loss function

Comparison of performances of L1 and L2 loss functions with and without outliers in a dataset.

Normal/Gaussian Distributions

Statistical Properties of Normal/Gaussian Distribution and why it is one of the most commonly used probability distribution in statistics.

Electricity Demand Analysis and Appliance Detection

Analysis of Electricity demand from a house on a time-series dataset. An appliance detection systems is also created using K-Means Clustering based on the electricity demand.