Ever since AlexNet was introduced in 2012, neural net research landscape fundamentally changed. With deep computational graphs, the possibilities are endless. We quickly iterated through a dizzying number of architectures such as AlexNet, VGGNe…

This is a continuation from the first post on exploring alternative neural computational models. To recap quickly, a neuron is fundamentally acting as a pattern matching unit by computing the dot product \(W \cdot X\…

Despite different variations of neural networks architectures, the computational model of a neuron hasn’t changed since inception. Every neuron has \(n\) incoming inputs \(X = (x_{1}, x_{2}, …, x_{n})\) with…

A while ago, kaggle hosted the ultrasound nerve segmentation challenge, which requires partipants to predict the nerve area (brachial plexus) in a given Ultrasound image. The task is to predict the segmentation mask for the…

You are to inflict maximum damage to the zerg army. There are two types of units - Warrior and Shield. Warriors do damage every second, while a shield protects your entire army for one second.…

Typically, a genetic algorithm works by the notion of maximizing the fitness. Consider a function \(y = x\), which is to be minimized in the interval \([-5, 5]\). One approach is use \(\frac{1}{x}\) as…

Given an integer \(N\), we need to find the number of integral pairs \((x, y) \mid x^2 + y^2 = N\), \(N \in [0, 2147483647]\), with a maximum of 100 such numbers in a file.…

Last week, I happened to read about Kleiber’s law while browsing through literature on natural evolution. The implications are really fascinating. It establishes a relationship between mass and metabolism as: \(metabolism = mass^{\frac{…

Today, I happened to type some random text and noticed interesting. This is a random sample of what i typed: gh jgh jg hjg hjg hjg hjg hjg hkg hjg hg jg hjg hjg hg…