Bayes’ Theorem is a way to calculate conditional probability. The formula is very simple to calculate, but it can be challenging to fit the right pieces into the puzzle. The first challenge comes from defining your event (A) and test (B); The second challenge is rephrasing your question so that you can work backwards: turning P(A|B) into P(B|A). The following image shows a…Continue
Added by Stephanie Glen on April 12, 2019 at 6:30am — No Comments
A non-technical look at A/B testing, based on Dan Siroker & Pete Koomen's book, A / B Testing, The Most Powerful Way to Turn Clicks Into Customers.
Perhaps the two most important points:
Added by Stephanie Glen on April 3, 2019 at 4:30pm — No Comments
Ensemble methods take several machine learning techniques and combine them into one predictive model. It is a two step process:
Added by Stephanie Glen on March 27, 2019 at 3:30pm — No Comments
SVMs (Support Vector Machines) are a way to classify data by finding the optimal plane or hyperplane that separates the data. In 2D, the separation is a plane; In higher dimensions, it's a hyperplane. For simplicity, the following picture shows how SVM works for a two-dimensional set.
Click on picture to zoom…Continue
Logistic regression is regressing data to a line (i.e. finding an average of sorts) so you can fit data to a particular equation and make predictions for your data. This type of regression is a good choice when modeling binary variables, which happen frequently in real life (e.g. work or don't work, marry or don't marry, buy a house or rent...). The logistic regression model is…Continue
Added by Stephanie Glen on March 22, 2019 at 11:30am — No Comments
This is a simple overview of the k-NN process. Perhaps the most challenging step is finding a k that's "just right". The square root of n can put you in the ballpark, but ideally you should use a training set (i.e. a nicely categorized set) to find a "k" that works for your data. Remove a few categorized data points and make them "unknowns", testing a few values for k to see what works.…Continue
Determining sample sizes is a challenging undertaking. For simplicity, I've limited this picture to the one of the most common testing situation: testing for differences in means. Some assumptions have been made (for example, normality and…Continue
The EM algorithm finds maximum-likelihood estimates for model parameters when you have incomplete data. The "E-Step" finds probabilities for the assignment of data points, based on a set of hypothesized probability…Continue
Added by Stephanie Glen on March 9, 2019 at 9:00am — No Comments
There are dozens of different hypothesis tests, so choosing one can be a little overwhelming. The good news is that one of the more popular tests will usually do the trick--unless you have unusual data or are working within very specific guidelines (i.e. in medical research). The following picture shows several tests for a single population, and what…Continue
Added by Stephanie Glen on March 7, 2019 at 7:30am — No Comments
In the nascent field of Data Science, myths are abound. Here's my top 10, scoured from the internet (where better than to find a myth or two?).
This one is only part myth. Historically, women have been discouraged from entering the computing sciences for many reasons unrelated to talent (see my previous post,…Continue
Whether you can call yourself a data scientist if you can't code is as hotly debated as Brexit. Type the question "Can you be a Data Scientist without coding?" into Google and you'll get a hundred different answers. The opinion will vary wildly depending on whether the author is a coder, or a non-coder. Search the job listings, and you won't find a definitive answer there either. A Glassdoor survey on the …Continue
Grab a copy of The Elements of Statistical Learning ("the machine learning bible") and you might be a little overwhelmed by the mathematics. For example, this equation (p.34), for a cubic smoothing spline, might send shivers down your spine if math isn't your forte:…Continue
I'm a female mathematician, statistician, and data scientist. How did I get to be all of those things? I wish I could say "I did well in school" or "I always loved computers." But the reality is, in middle/high school, I was bored stiff. I excelled in art, skipping classes, and bombing exams. At 16, I dropped out of school to begin an illustrious career in office cleaning.
The odds were stacked against me entering the computing industry for many…Continue
Logistic regression (LR) models estimate the probability of a binary response, based on one or more predictor variables. Unlike linear regression models, the dependent variables are categorical. LR has become very popular, perhaps because of the wide availability of the procedure in software. Although LR is a good choice for many situations, it doesn't work well for all situations. For example:
Added by Stephanie Glen on February 2, 2019 at 6:55am — No Comments