A python gist for balancing (re-sampling) a training dataset to match a reported positive-negative class ratio, in the unseen dataset When we know the unseen’s pos-neg class ratio (or “guess” it from the LB..) we should give a try at balancing the training dataset, to reflect it. I wrote a python gist for it, using […]

# Category archives for **Machine Learning**

As I mentioned in my previous post, a linear decision boundary is rarely enough to describe a real life decision boundary. The Idea behind a non-linear logistic regression algorithms is to “grasp” the more complex mathematical relationship between our set of X features and their respective y label. One way is finding a polynomial relation, also known […]

Let’s try to predict our chances of getting admitted to Msc studies, based on our Bsc degree GPA and years of experience. Our csv should look like: 2 “X” columns: Bsc-GPA and experience and a “y” column of if-admitted value (0 / 1). We can simulate some data with excel-like functions (I’m using LibreOffice calc on […]

Let’s test drive simdfied library with a linear regression example. We’ll use MLplaygroung.org, that uses simdfied for Machine Learning and can read csv or mat files. For example, a csv file representing house prices according to its square-foot and number of bedrooms: square foot, #bedrooms, price 2461.68 , 4 , 467883 1872, 4 , 385983 And […]

Lately I’ve been working on simdfied – a Machine Learning javascript library, utilizing SIMD! Earlier this year, Intel announced their collaboration with Google and Mozilla, enabling a preview native SIMD support for firefox-nightly and chromium browsers; they provided some really nice stats for gaming implementations. For me it was time for an idea I’ve had for a while; implementing a javascript […]

With a cross-giants collaboration between Intel and the leading modern browsers Mozilla and Google-Chrome, a native javascript SIMD performance is now available! In his article, Bringing SIMD to JavaScript, Mohammad Reza Haghighat presents interesting benchmarks running complicated visualizations. I’m much more interested with the ML implications of using SIMD over javascript. I believe this joint effort should […]