By Mladen Jovanovic on 20/01/2015
Confidence intervals gives us the range of a given statistic when generalizing from a sample to a population. The simplest example could be mean of a sample (e.g. average height) – what we are interested in are the generalizations (or inferences) from this sample to a population (e.g. average height in population). Due the sampling error we are not…
By Mladen Jovanovic on 06/11/2014
We collect more and more data and it is becoming increasingly difficult to make meaning out of it. What I would like to do is to present one simple way to make the meaning out of session GPS data using LOF and Clustering. Most GPS units produce multiple features p compared to number of observations n, so we are…
By Mladen Jovanovic on 01/11/2014
I have recently read two great papers on using alternative metrics when analyzing countermovement jump (CMJ) with the goal of evaluating both acute (neuromuscular fatigue, or NMF) and chronic training effects written by Rob Gathercole et al.
I was amazed by how much new food for thought have been inside and how great was the novel combination of inferential statistics…
By Mladen Jovanovic on 22/10/2014
Here is the another ‘playbook’, but this time on my own data set during high-frequency project I did in 2013. The data set features estimated 1RM using velocity (which I measure during the lifts). I have also measure Peak Velocity and Mean Power in CMJ w/20kg before lower body workouts. Those four are response variables.
By Mladen Jovanovic on 18/10/2014
In the previous part I’ve introduced multivariate modeling of impulse and response using Banister model. In this part I will continue with exploration of this model, mainly visualizing reaction predicted by the model on standardize impulse (load) and compare prediction using multiple impulses. I will use same data sets: one by Skiba and one randomly generated as in first…
By Mladen Jovanovic on 14/10/2014
Banister Impulse~Response Model in R Before you start reading this post, please read EXCELLENT paper by Clark and Skiba, especially on the topic of Banister impulse-response model. I decided to write code in R, but also allow for multivariate analysis (where impulse can be multiple variables, as is the case in sports) which can speed the thing...
By Mladen Jovanovic on 28/09/2014
I have just started reading about machine learning and predictive analytics in general, so I wanted to share my planned journey for those who wish to start the same.
By Mladen Jovanovic on 24/09/2014
In the previous post I’ve shared the novel idea on how to ‘aggregate’ wellness categories into positive/negative wellness score.
By Mladen Jovanovic on 16/09/2014
New Way to Calculate Wellness From Questionnaires In the video below I go over the usual problems of calculating total Wellness from different categories and present the new way to calculate total wellness score that is more sensitive.
By Mladen Jovanovic on 12/09/2014
Statistics 101 – Introduction to ‘Model Thinking’ I am just reading Data Science for Business book, which is by the way great, and finishing Discovering Statistics Using R. Both of these provide interesting viewpoint on building models. In most other stats books, linear regression and correlation are explained by trying to find linear relationship between variables – using ‘model...