Machine Learning Strategy

In modern big era, machine learning strategy plays significant role in ultimate fate of a machine learning project. Let's say we have trained a classifier with 90% accuracy on test examples. But that accuracy is not be good enough for the application. To improve classifier’s accuracy, we can try a ra... read more

Data Science 101: simplest possible hands-on example?

Dr. Rizz I have been thinking about writing up a simplest possible data science example that people can play with, and it gives a taste of predicting something useful. In today’s hot real estate market, I choose the example of freely and openly available Boston Housing dataset available at UC... read more

Hyperparameter Tuning of Deep Learning Algorithm

A deep learning algorithm requires tuning of following possible set of parameters (also known as hyperparameters): $\alpha$ - learning rate - the most important hyperparameter to tune learning rate decay - probably second in importance mini-batch size - as important as learning rate decacy mini-batch iterations number of layers $\beta$ ... read more

407: she loves me; she loves me not?

Commuters of Greater Toronto Area (GTA) are plagued by the question of whether to take 407 in their daily commute, or not. People come up with different heuristics from listening to traffic news, anticipating good/bad weather, guessing traffic conditions on a particular day of the week to avoiding rush hours, ... read more

Gradient Descent with Momentum

Gradient descent is commonly used search algorithm in machine learning. We use it to learn parameters of prediction model. One of the problems with straightforward gradient descent is that it oscillates too much on vertical axis (as show in above image). This oscillations slows down the convergence to minimum (the ... read more

More blog posts